Olena-patches

olena-patches@lrde.epita.fr

9625 discussions

last-svn-commit-199-g26fd960 Group Millet2008 documentation files.

by green

* green/doc/annotating: New directory. * green/doc/annotating/class.txt: New hierarchy description. * green/doc/annotating/testMillet2008.txt: New result interpretations. * green/doc/annotating/syntheseMillet2008.txt: New feature extractions. * green/demo/annotating/class.txt: Remove this obsolete file. --- milena/sandbox/ChangeLog | 10 + milena/sandbox/green/demo/annotating/class.txt | 33 --- milena/sandbox/green/doc/annotating/class.txt | 42 +++ .../green/doc/annotating/syntheseMillet2008.txt | 283 ++++++++++++++++++++ milena/sandbox/green/doc/annotating/testMillet2008 | 69 +++++ 5 files changed, 404 insertions(+), 33 deletions(-) delete mode 100644 milena/sandbox/green/demo/annotating/class.txt create mode 100644 milena/sandbox/green/doc/annotating/class.txt create mode 100644 milena/sandbox/green/doc/annotating/syntheseMillet2008.txt create mode 100644 milena/sandbox/green/doc/annotating/testMillet2008 diff --git a/milena/sandbox/ChangeLog b/milena/sandbox/ChangeLog index 6435445..a096559 100644 --- a/milena/sandbox/ChangeLog +++ b/milena/sandbox/ChangeLog @@ -75,6 +75,16 @@ * bin/pgm_to_pbm.cc, * bin/ppm_negate.cc: New. +2009-12-16 Yann Jacquelet <jacquelet(a)lrde.epita.fr> + + Group Millet2008 documentation files. + + * green/doc/annotating: New directory. + * green/doc/annotating/class.txt: New hierarchy description. + * green/doc/annotating/testMillet2008.txt: New result interpretations. + * green/doc/annotating/syntheseMillet2008.txt: New feature extractions. + * green/demo/annotating/class.txt: Remove this obsolete file. + 2009-12-16 Yann Jacquelet <jacquelet(a)lrde.epita.fr> Correct typo into directories. diff --git a/milena/sandbox/green/demo/annotating/class.txt b/milena/sandbox/green/demo/annotating/class.txt deleted file mode 100644 index ec7e92e..0000000 --- a/milena/sandbox/green/demo/annotating/class.txt +++ /dev/null @@ -1,33 +0,0 @@ -Dans scribo les classes de documents sont: - -- image seule (photo) -- courrier manuscrit -- courrier dactylographié -- facture ou feuille organisée sous forme de tableau -- article de journal -- document composite -- infographie (logo, carte, dessin vectoriel) -- slide (copie d'écran ou sauvegarde au format image) - - -Où se situent les faxes ? - - -Description d'un fax: -- ratio l/L assimilable à du A4 ou US Letter -- nombre de couleurs très petit, < 1000 -- plutôt du N & B, mais quelques faxes sont en couleurs -- généralement un % élevé de blanc (la page vierge) - - -Description d'un article de journal: -- nombre élevés de couleurs (> 10 000) en général -- quelques articles de conférences autour de 5000 - - -Description des photos: -- un nombre très élevé de couleurs -- ratio non A4, non US Letter - - - diff --git a/milena/sandbox/green/doc/annotating/class.txt b/milena/sandbox/green/doc/annotating/class.txt new file mode 100644 index 0000000..0fa5cf9 --- /dev/null +++ b/milena/sandbox/green/doc/annotating/class.txt @@ -0,0 +1,42 @@ +Dans scribo les classes de documents sont: + +- image seule (photo) +- courrier manuscrit +- courrier dactylographié +- facture ou feuille organisée sous forme de tableau +- article de journal +- document composite +- infographie (logo, carte, dessin vectoriel) +- slide (copie d'écran ou sauvegarde au format image) + +Le problème avec cette hiérarchie c'est quelle induit une connaissance +fonctionnelle des images traitées et que les extracteurs de primitives font +plutôt ressortir des informations relatives à la nature de l'image. + +Par exemple, les slides peuvent contenir aussi bien du texte, que des photos +ou des cliparts. C'est une forme de "composite" par définition. + +Où se situent les faxes ? + + +Description d'un fax: +- ratio l/L assimilable à du A4 ou US Letter +- nombre de couleurs très petit, < 1000 +- plutôt du N & B, mais quelques faxes sont en couleurs +- généralement un % élevé de blanc (la page vierge) + + +Description d'un article de journal: +- nombre élevés de couleurs (> 10 000) en général +- quelques articles de conférences autour de 5000 + + +Description des photos: +- un nombre très élevé de couleurs +- ratio non A4, non US Letter + + +Google hiérarchie: +- line drawing +- clipart +- photo diff --git a/milena/sandbox/green/doc/annotating/syntheseMillet2008.txt b/milena/sandbox/green/doc/annotating/syntheseMillet2008.txt new file mode 100644 index 0000000..b7b3648 --- /dev/null +++ b/milena/sandbox/green/doc/annotating/syntheseMillet2008.txt @@ -0,0 +1,283 @@ + +Millet utilise un certain nombre de descripteurs (p 65, Chapitre 3) +pour l'annotation automatique. Son objectif est de formuler une +annotation automatique simple à partir de descripteurs de bas niveaux +qui permettent de réaliser l'opération en moins de 10 secondes. + +L'auteur présente dans le chapitre 4, p 77 l'arbre de classification +et l'utilisation précise qu'il fait de ces descripteurs. + + +Reconnaissance des images de clipart versus photographies +Reconnaissance des images Noir & Blanc versus Couleur +Reconnaissance Interieur versus Exterieur +Autre reconnaissance cartes, photos, peintures. + + + /---------Image------------\ + | | + v v + /-------+------+-------\ /-------+----------\ + | | | | | | | + v v | v v v v +Carte Clipart | Peinture Couleur N&B N&B colorisé + | + v + /----Photo----------\ + | | + v | + /----------\ | + | | | + v | v +Intérieur | Visages + | + v + /-----Extérieur------\ + | | + v v + /-----\ /-----\ + | | | | + v v v v +Jour Nuit Nature Urbain + + +Note: +----- +Lorsque d'un concept, on part de sa droite et de sa gauche, il s'agit en fait +de tests différents effectués en parallèles (ex.: Image, Photo, Extérieur). + +Lorsqu'on aboutit à une bifurcation et qu'il ne s'agit pas d'un concept, cela +correspond à un seul test dont les résultats sont répartis dans des classes +différentes. + +Fax +Magazine + + + +ANNEXE 1 : SYNTHESE DESCRIPTEURS MILLET +======================================= + +* RGB-64 +-------- +Chaque composante est quantifiée sur 4 valeurs (4x4x4 = 64), puis on construit +l'histogramme à 64 composantes. + + +* TSVal +------- +A partir de l'espace RGB d'origine, on effectue +la transformation. Attention les composantes ne sont pas quantifiées +de la même manière (18 valeurs pour la teinte [17 + 1 pour le +négatif], Val et S sont quantifiées sur 3 valeurs +chacunes. L'histogramme obtenu dans cet espace contiendra 162 +composantes. + +Val = max(R,G,B) +S = [max(R,G,B) - min(R,G,B)] / max(R,G,B) + +Si R = max(R,G,B) +Alors T = 60 * (G-B) / [max(R,G,B) - min(R,G,B)] + +Si G = max(R,G,B) +Alors T = 60 * (2 + (B-R) / [max(R,G,B) - min(R,G,B)]) + +Si B = max(R,G,B) +Alors T = 60 * (4 + (R-G) / [max(R,G,B) - min(R,G,B)]) + + +* RGB-64-9 +---------- +L'image est décomposée en 9 sous images identiques, puis on applique le +descripteur RGB-64. L'histogramme obtenu est la concaténation des 9 +histogrammes soit un histogramme résultant à 576 composantes. + + +* BIC +----- +L'image est quantifiée en 216 couleurs (RGB-216) 6 valeurs par couleur. On +distingue ensuite les pixels du bord des pixels intérieurs. Les pixels +intérieurs ont la même couleur que leurs 4-voisins. Un histogramme est crée +pour chaque groupe de pixels. Puis on concatène +les histogrammes en un seul de 432 composantes. + + +* LEP +----- +On construit une image de gradient avec un sobel 3x3, puis cette image est +seuillée à 100 et devient donc binaire. On étiquette les configurations +obtenues avec un masque 3x3 classique (convolution). Il existe 2^9 +configurations possibles. On construit alors un histogramme à 512 composantes. + +masque de convolution pour l'identification de la configuration: +001 002 004 +008 256 016 +032 064 128 + + +* GABOR +------- +Très coûteux, la réponse de 24 filtres sur une image 300x300 est de l'ordre de + 1s. On utilise ici des filtres de taille 8x8. + +F(x,y) = [1 / (2 PI sigma_x sigma_y)] + * exp[(-1/2)(x^2/sigma_x^2 + y^2/sigma_y^2)] * exp[j2 PI Wx] + +sigma_x et sigma_y sont des paramètres d'échelles. +W permet de changer l'orientation. + +Il existe une formule pour trouver les coefficients des filtres une fois que +l'on a déterminé le nombre d'orientations et d'échelles utilisées. On ne +retiendra, en sortie de filtre, que la variance et l'énergie. L'auteur +préconise d'utiliser 6 directions et 4 échelles. + + +--------------GABOR---------MATLAB---------------------------------------------- +function gb=gabor_fn(sigma,theta,lambda,psi,gamma) + +sigma_x = sigma; +sigma_y = sigma/gamma; + +% Bounding box +nstds = 3; +xmax = max(abs(nstds*sigma_x*cos(theta)),abs(nstds*sigma_y*sin(theta))); +xmax = ceil(max(1,xmax)); +ymax = max(abs(nstds*sigma_x*sin(theta)),abs(nstds*sigma_y*cos(theta))); +ymax = ceil(max(1,ymax)); +xmin = -xmax; ymin = -ymax; +[x,y] = meshgrid(xmin:xmax,ymin:ymax); + +% Rotation +x_theta=x*cos(theta)+y*sin(theta); +y_theta=-x*sin(theta)+y*cos(theta); + +gb=exp(-.5*(x_theta.^2/sigma_x^2+y_theta.^2/sigma_y^2)).*cos(2*pi/lambda*x_theta+psi); + +http://matlabserver.cs.rug.nl/cgi-bin/matweb.exe +Gabor Filters. Tech. rep., 2002. +--------------GABOR---------MATLAB---------------------------------------------- + + +* PROJECTION +------------ +L'image est redimentionnée en 100x100. Un coup de Sobel 3x3, puis une +binarisation avec s=100. On considère les moitiées verticales puis +horizontales. Pour les moitiées verticales, on réalise la somme sur chacune +des lignes, on obtient alors un vecteur de taille 100. +Pour les moitiées horizontales, on réalise les sommes des colonnes. On obtient +alors un vecteur taille 100. Les 4 sous vecteurs obtenus peuvent être +concaténés pour former un unique descipteur de 400 composantes. + + + +ANNEXE 2 : SYNTHESE CLASSIFICATION MILLET +========================================= + + +* Reconnaissance des images de clipart versus photographies +----------------------------------------------------------- + +Construire l'histogramme de l'image, éliminer toutes les couleurs peu +représentées et ensuite les comptées. Moins de 50 couleurs / 256, c'est un +clipart, plus de 150 / 256 c'est une photographie. Entre les deux, indécision. +D'après l'auteur, le test n'est pas complètement fiable et peu être évincé, tout +dépend de la qualité de la compression des cliparts. + +Diviser l'image en 16 imagettes (4x4), calculer la suite pour chaque imagette. +Trouver le maxima de l'histogramme (noté p)/ p = argmax H(x), x in [0,255]. +Puis on calcule un sigma autour du pic. + +sigma^2 = Sum(p-1,p-5) r(x) pour p > 250 +sigma^2 = Sum(p+1,p+5) r(x) pour p < 005 +sigma^2 = (1/2) (Sum(p-1,p-5) r(x) + Sum(p+1,p+5) r(x)) pour les autres p + +r(x) = ((H(x)/H(p)) * (x-p))^2 + + +Le descripteur de l'image entière est alors le maximum des 16 sigma^2. Si le +pic est resseré, c'est plutôt un clipart (< 5), sinon plutôt une photographie +(> 40). Le seuil est fixé à 15. + + + +* Reconnaissance des images Noir & Blanc versus Couleur +------------------------------------------------------- + +Achromatisme (en passant par TSVal) + +si |R - G| < s +et si |R - B| < s alors R ~ G ~ B +et si |G - B| < s + +==> T = -1 +==> Val = (R + G + B)/3 + +L'auteur propose s = 0,03 + +Une image est dite Noir & Blanc (donc Achromatique) si + de 99% des pixels le +sont. Le test peut être appliqué aux images étiquetées comme des cliparts, ou +des photographies. + + +Faible saturation: + +On établit l'histogramme de la saturation de l'image (0 à 255). Si plus de 95% +des pixels sont inférieurs à une saturation de 100, alors l'image est faiblement +saturée. Ce test ne s'applique sur les photographies, surtout pas aux cliparts. + + +Dominance de la teinte (image jaunie): + +On construit l'histogramme de la teinte (0 à 255). On localise le maximum de +l'histogramme. Si + de 95% des pixels ont une distance < 20 du maximum, alors +l'image est considérée comme Noir & Blanc et on peut donner sa teinte dominante. +De plus il faut qu'au moins 50% des pixels soient faiblement saturés. + +T = -1 Achromatisme + 0 <= T < 14 Rouge + 14 <= T < 29 Orange + 29 <= T < 45 Jaune + 45 <= T < 113 Vert +113 <= T < 149 Cyan +149 <= T < 205 Bleu +205 <= T < 235 Violet +235 <= T < 242 Rose, Magenta +242 <= T < 255 Rouge + + + 0 <= V < 82 Noir + 82 <= V < 179 Gris +179 <= V < 255 Blanc + + +Pour un pixel (S, Val) +si |184 - S| + |65 - Val| < |255 - S| + |125 - Val| alors marron sinon orange. + +Et encore +si la teinte est jaune et Val < 80 alors vert sinon jaune. + + + +* Reconnaissance Interieur versus Exterieur +------------------------------------------- + +Utilisation des descripteurs LEP et RGB-64. Expérience avec un kpp ~ 80%. Puis +expérimentation avec un SVM ~ 90% de réussite. Finalement, inclusion dans le SVM +de LEP, RGB-64, TSVal-125, projection, Bic, Gabor. + + + + +* Localisation des visages +-------------------------- + +Une semaine d'apprentissage, 5 secondes pour leur localisation. Je ne suis pas +sûr que cette méthode soit pertinente pour notre contexte applicatif. + + + +* Autre classification +---------------------- + +Utilisation de SVM et de tous les descripteurs possibles comme pour Interieur +versus Exterieur. diff --git a/milena/sandbox/green/doc/annotating/testMillet2008 b/milena/sandbox/green/doc/annotating/testMillet2008 new file mode 100644 index 0000000..a5cf22d --- /dev/null +++ b/milena/sandbox/green/doc/annotating/testMillet2008 @@ -0,0 +1,69 @@ +Dans le cadre de nos recherches, nous testons le classifieur de +Millet2008 pour la classification des images en clipart ou en +photographies. Deux tests sont proposés, le premier est basé sur le +nombre de couleurs de l'image après un seuillage de l'histogramme de +façon à éliminer les couleurs de faible contribution; le second test +consiste à extraire le pic de l'histogramme et à mesurer une +"variance" normalisée autour de ce pic. Le premier test part du +principe que les photographies possèdent de très nombreuses couleurs, +tandis que les cliparts n'en ont que quelques unes. Le test n'est pas +toujours vérifié. Le second test s'intéresse à la nature de +l'histogramme. L'idée étant de montrer qu'un histogramme continu, donc +avec une "variance" élevé autour du pic principal est plutôt +représentatif des phtotographies, alors qu'une "variance" faible tient +d'avantage du clipart (peu de couleurs, couleurs isolées). Selon +l'auteur, ce second test donne de très bon résultats. Il peut être +affiné en divisant l'image en 16 sous parties et en analysant chacune +des imagettes avec la technique décrite ci-dessus. Dans ce cas, la +"variance" de l'image est le max des "variances" des imagettes. Les +repères suivnt sont donnés par l'auteur, "variance" < 5 pour les +clipart et "variance" ~ 40 pour les photographies (seuil fixé à 15). + + +Test A (Nombre de couleurs dans l'histogramme): +----------------------------------------------- + + +Test B ("Variance" autour du pic principal): +-------------------------------------------- + +bill: Les images de factures contiennent un ensemble d'informations +groupés sous forme de tableau. Il peut y avoir plusieurs tableaux, +visibles ou non. Pour la pupart des histogrammes observés, +l'information en niveau de gris montre un grand pic blanc sans +voisinage ou avec un voisinage de taille insignifiante. Toutes les +factures devraient être classées comme des cliparts sauf exception. + +deux images sortent du lot: +- bill17.ppm => 12.7828 +- bill23.ppm => 17.2565 + +bill17 contient des dégradés qui viennent épaissir le pic blanc, du +coup le voisinage autour du pic devient plus important. + +bill23 contient un logo et une image de fond en transparence sur le +tableau. Par ailleurs, certains textes sont sur un fond +composite. Tous ces éléments viennent épaissir le pic blanc. + +Résultat : 29 images / 30 passent le test + + +photo: Les images de l'AFP ont la particularité d'être cadré serré +pour faire ressortir un élément clé de la scène photographiée. On +s'attend, d'après l'auteur à obtenir une "variance" proche de 40 (seuil ~ 15). + +5 vrai positifs: +- photo01 ==> 42.9224 +- photo10 ==> 48.0298 +- photo25 ==> 33.6833 +- photo27 ==> 42.9025 +- photo28 ==> 43.5689 + +des scores faibles: +- + + + + +Test C ("Variance" sur les 16 imagettes): +----------------------------------------- \ No newline at end of file -- 1.5.6.5

15 years

last-svn-commit-200-ge9bf3ef Introduce the local deviation operator (Millet2008) which subdivise the image in 16 parts.

by green

* green/demo/annotating/stddev_color_16: New directory. * green/demo/annotating/stddev_color_16/Makefile.am: New Makefile. * green/demo/annotating/stddev_color_16/stddev_color_16.cc: New demo. --- milena/sandbox/ChangeLog | 9 + .../{bic => stddev_color_16}/Makefile.am | 0 .../annotating/stddev_color_16/stddev_color_16.cc | 178 ++++++++++++++++++++ 3 files changed, 187 insertions(+), 0 deletions(-) copy milena/sandbox/green/demo/annotating/{bic => stddev_color_16}/Makefile.am (100%) create mode 100644 milena/sandbox/green/demo/annotating/stddev_color_16/stddev_color_16.cc diff --git a/milena/sandbox/ChangeLog b/milena/sandbox/ChangeLog index a096559..423c61b 100644 --- a/milena/sandbox/ChangeLog +++ b/milena/sandbox/ChangeLog @@ -77,6 +77,15 @@ 2009-12-16 Yann Jacquelet <jacquelet(a)lrde.epita.fr> + Introduce the local deviation operator (Millet2008) which + subdivise the image in 16 parts. + + * green/demo/annotating/stddev_color_16: New directory. + * green/demo/annotating/stddev_color_16/Makefile.am: New Makefile. + * green/demo/annotating/stddev_color_16/stddev_color_16.cc: New demo. + +2009-12-16 Yann Jacquelet <jacquelet(a)lrde.epita.fr> + Group Millet2008 documentation files. * green/doc/annotating: New directory. diff --git a/milena/sandbox/green/demo/annotating/bic/Makefile.am b/milena/sandbox/green/demo/annotating/stddev_color_16/Makefile.am similarity index 100% copy from milena/sandbox/green/demo/annotating/bic/Makefile.am copy to milena/sandbox/green/demo/annotating/stddev_color_16/Makefile.am diff --git a/milena/sandbox/green/demo/annotating/stddev_color_16/stddev_color_16.cc b/milena/sandbox/green/demo/annotating/stddev_color_16/stddev_color_16.cc new file mode 100644 index 0000000..5e95c76 --- /dev/null +++ b/milena/sandbox/green/demo/annotating/stddev_color_16/stddev_color_16.cc @@ -0,0 +1,178 @@ +// COMPUTING THE STDEV OF THE COLORS + +#include <iostream> +#include <sstream> + +#include <mln/img_path.hh> + +#include <mln/accu/max_site.hh> +#include <mln/accu/stat/histo1d.hh> + +#include <mln/core/macros.hh> +#include <mln/core/image/image1d.hh> +#include <mln/core/image/image2d.hh> +#include <mln/core/image/image3d.hh> + +#include <mln/debug/println.hh> + +#include <mln/data/compute.hh> +#include <mln/data/fill.hh> +#include <mln/data/transform.hh> +#include <mln/data/paste.hh> + +#include <mln/fun/v2v/rgb8_to_int_u8.hh> + +#include <mln/io/ppm/load.hh> +#include <mln/io/pgm/save.hh> +#include <mln/io/plot/save_image_sh.hh> + +#include <mln/math/sqr.hh> + +#include <mln/opt/at.hh> + +#include <mln/value/rgb8.hh> +#include <mln/value/int_u.hh> + + +float r(short p, unsigned histo_p, short x, unsigned histo_x) +{ + float result = mln::math::sqr(((float)histo_x / histo_p) * (x-p)); + + return result; +} + +unsigned stddev_color(mln::image2d<mln::value::int_u8> input_int_u8, + const char *name_histo, + const char *name_image) +{ + typedef mln::point1d t_point1d; + typedef mln::value::rgb8 t_rgb8; + typedef mln::value::int_u8 t_int_u8; + typedef mln::image2d<t_rgb8> t_image2d_rgb8; + typedef mln::image2d<t_int_u8> t_image2d_int_u8; + typedef mln::image1d<unsigned> t_histo1d; + typedef mln::fun::v2v::rgb8_to_int_u8 t_rgb8_to_int_u8; + typedef mln::accu::meta::stat::histo1d t_histo1d_fun; + typedef mln::accu::max_site<t_histo1d> t_max_site_fun; + + t_histo1d histo; + + std::cout << "histo : " << name_histo << std::endl; + std::cout << "image : " << name_image << std::endl; + + histo = mln::data::compute(t_histo1d_fun(), input_int_u8); + + mln::io::pgm::save(input_int_u8, name_image); + mln::io::plot::save_image_sh(histo, name_histo); + mln::debug::println(histo); + + // Find the peak of the histogram + unsigned v_max = mln::opt::at(histo, 0); + short p_max = 0; + + mln_piter_(t_histo1d) p(histo.domain()); + + for_all(p) + { + if (v_max < histo(p)) + { + v_max = histo(p); + p_max = p.ind(); + } + } + + // Compute the specific stddev + + float stddev_low = 0.0; + float stddev_up = 0.0; + float stddev = 0.0; + + if (250 > p_max) + for (short i = p_max+1; i < p_max+6; ++i) + stddev_up += r(p_max, mln::opt::at(histo,p_max), + i, mln::opt::at(histo,i)); + + if (5 < p_max) + for (short i = p_max-1; i > p_max-6; --i) + stddev_low += r(p_max, mln::opt::at(histo,p_max), + i, mln::opt::at(histo,i)); + + stddev = (250 < p_max)? stddev_low : (5 > p_max)? stddev_up : + (stddev_low + stddev_up)/2; + + std::cout << "max_site : " << p_max << std::endl; + std::cout << "h(max_site) : " << v_max << std::endl; + std::cout << "stddev_up : " << stddev_up << std::endl; + std::cout << "stddev_low : " << stddev_low << std::endl; + std::cout << "stddev : " << stddev << std::endl; + + return 0; +} + +unsigned stddev_color_16(const std::string& image) +{ + typedef mln::point1d t_point1d; + typedef mln::value::rgb8 t_rgb8; + typedef mln::value::int_u8 t_int_u8; + typedef mln::image2d<t_rgb8> t_image2d_rgb8; + typedef mln::image2d<t_int_u8> t_image2d_int_u8; + typedef mln::image1d<unsigned> t_histo1d; + typedef mln::fun::v2v::rgb8_to_int_u8 t_rgb8_to_int_u8; + typedef mln::accu::meta::stat::histo1d t_histo1d_fun; + typedef mln::accu::max_site<t_histo1d> t_max_site_fun; + + t_image2d_rgb8 input_rgb8; + t_image2d_int_u8 input_int_u8; + + mln::io::ppm::load(input_rgb8, image.c_str()); + input_int_u8 = mln::data::transform(input_rgb8, t_rgb8_to_int_u8()); + + // IMAGE SPLITTING PHASE + mln::box2d domain = input_int_u8.domain(); + mln::point2d pmin = domain.pmin(); + mln::point2d pmax = domain.pmax(); + + unsigned sz_row = (pmax.row() - pmin.row())/ 4; + unsigned sz_col = (pmax.col() - pmin.col())/ 4; + + std::cout << domain << std::endl; + + // Divide the domain in nine sub-domains. + for (unsigned i = 0; i < 4; ++i) + for (unsigned j = 0; j < 4; ++j) + { + mln::point2d min(pmin.row()+sz_row*i,pmin.col()+sz_col*j); + mln::point2d max(pmin.row()+sz_row*(i+1),pmin.col()+sz_col*(j+1)); + mln::box2d dom(min,max); + + std::cout << dom << std::endl; + + // Save it + t_image2d_int_u8 input_1o16_int_u8(dom); + std::ostringstream name_histo(""); + std::ostringstream name_image(""); + + name_histo << "histo" << i << "_" << j << ".sh"; + name_image << "image" << i << "_" << j << ".ppm"; + + mln::data::paste(input_int_u8 | dom, input_1o16_int_u8); + + stddev_color(input_1o16_int_u8, + name_histo.str().c_str(), + name_image.str().c_str()); + } + + return 0; +} + + +int main() +{ +// unsigned val = stddev_color_16(ANNOTATING_PHOTO_IMG_PATH "/photo01.ppm"); + unsigned val = stddev_color_16(ANNOTATING_LOGO_IMG_PATH "/logo06.ppm"); + +// std::cout << "nb color : " << val << std::endl; + + return 0; +} + -- 1.5.6.5

15 years

last-svn-commit-201-g07b0d84 Test the local deviation operator (Millet2008) on the annotating base.

by green

* green/exp/annotating/stddev_color_16: New directory. * green/exp/annotating/stddev_color_16/Makefile.am: New Makefile. * green/exp/annotating/stddev_color_16/stddev_color_16.cc: New demo. --- milena/sandbox/ChangeLog | 8 ++ .../{nb_color => stddev_color_16}/Makefile.am | 0 .../annotating/stddev_color_16/stddev_color_16.cc | 87 ++++++++++++-------- 3 files changed, 60 insertions(+), 35 deletions(-) copy milena/sandbox/green/exp/annotating/{nb_color => stddev_color_16}/Makefile.am (100%) copy milena/sandbox/green/{demo => exp}/annotating/stddev_color_16/stddev_color_16.cc (70%) diff --git a/milena/sandbox/ChangeLog b/milena/sandbox/ChangeLog index 423c61b..7ec7a8a 100644 --- a/milena/sandbox/ChangeLog +++ b/milena/sandbox/ChangeLog @@ -77,6 +77,14 @@ 2009-12-16 Yann Jacquelet <jacquelet(a)lrde.epita.fr> + Test the local deviation operator (Millet2008) on the annotating base. + + * green/exp/annotating/stddev_color_16: New directory. + * green/exp/annotating/stddev_color_16/Makefile.am: New Makefile. + * green/exp/annotating/stddev_color_16/stddev_color_16.cc: New demo. + +2009-12-16 Yann Jacquelet <jacquelet(a)lrde.epita.fr> + Introduce the local deviation operator (Millet2008) which subdivise the image in 16 parts. diff --git a/milena/sandbox/green/exp/annotating/nb_color/Makefile.am b/milena/sandbox/green/exp/annotating/stddev_color_16/Makefile.am similarity index 100% copy from milena/sandbox/green/exp/annotating/nb_color/Makefile.am copy to milena/sandbox/green/exp/annotating/stddev_color_16/Makefile.am diff --git a/milena/sandbox/green/demo/annotating/stddev_color_16/stddev_color_16.cc b/milena/sandbox/green/exp/annotating/stddev_color_16/stddev_color_16.cc similarity index 70% copy from milena/sandbox/green/demo/annotating/stddev_color_16/stddev_color_16.cc copy to milena/sandbox/green/exp/annotating/stddev_color_16/stddev_color_16.cc index 5e95c76..c196595 100644 --- a/milena/sandbox/green/demo/annotating/stddev_color_16/stddev_color_16.cc +++ b/milena/sandbox/green/exp/annotating/stddev_color_16/stddev_color_16.cc @@ -2,6 +2,7 @@ #include <iostream> #include <sstream> +#include <boost/filesystem.hpp> #include <mln/img_path.hh> @@ -27,6 +28,7 @@ #include <mln/io/plot/save_image_sh.hh> #include <mln/math/sqr.hh> +#include <mln/math/max.hh> #include <mln/opt/at.hh> @@ -41,9 +43,7 @@ float r(short p, unsigned histo_p, short x, unsigned histo_x) return result; } -unsigned stddev_color(mln::image2d<mln::value::int_u8> input_int_u8, - const char *name_histo, - const char *name_image) +float stddev_color(mln::image2d<mln::value::int_u8> input_int_u8) { typedef mln::point1d t_point1d; typedef mln::value::rgb8 t_rgb8; @@ -57,15 +57,8 @@ unsigned stddev_color(mln::image2d<mln::value::int_u8> input_int_u8, t_histo1d histo; - std::cout << "histo : " << name_histo << std::endl; - std::cout << "image : " << name_image << std::endl; - histo = mln::data::compute(t_histo1d_fun(), input_int_u8); - mln::io::pgm::save(input_int_u8, name_image); - mln::io::plot::save_image_sh(histo, name_histo); - mln::debug::println(histo); - // Find the peak of the histogram unsigned v_max = mln::opt::at(histo, 0); short p_max = 0; @@ -100,16 +93,10 @@ unsigned stddev_color(mln::image2d<mln::value::int_u8> input_int_u8, stddev = (250 < p_max)? stddev_low : (5 > p_max)? stddev_up : (stddev_low + stddev_up)/2; - std::cout << "max_site : " << p_max << std::endl; - std::cout << "h(max_site) : " << v_max << std::endl; - std::cout << "stddev_up : " << stddev_up << std::endl; - std::cout << "stddev_low : " << stddev_low << std::endl; - std::cout << "stddev : " << stddev << std::endl; - - return 0; + return stddev; } -unsigned stddev_color_16(const std::string& image) +float stddev_color_16(const std::string& image) { typedef mln::point1d t_point1d; typedef mln::value::rgb8 t_rgb8; @@ -134,8 +121,7 @@ unsigned stddev_color_16(const std::string& image) unsigned sz_row = (pmax.row() - pmin.row())/ 4; unsigned sz_col = (pmax.col() - pmin.col())/ 4; - - std::cout << domain << std::endl; + float stddev = 0.0; // Divide the domain in nine sub-domains. for (unsigned i = 0; i < 4; ++i) @@ -145,34 +131,65 @@ unsigned stddev_color_16(const std::string& image) mln::point2d max(pmin.row()+sz_row*(i+1),pmin.col()+sz_col*(j+1)); mln::box2d dom(min,max); - std::cout << dom << std::endl; - // Save it t_image2d_int_u8 input_1o16_int_u8(dom); - std::ostringstream name_histo(""); - std::ostringstream name_image(""); - - name_histo << "histo" << i << "_" << j << ".sh"; - name_image << "image" << i << "_" << j << ".ppm"; mln::data::paste(input_int_u8 | dom, input_1o16_int_u8); - stddev_color(input_1o16_int_u8, - name_histo.str().c_str(), - name_image.str().c_str()); + stddev = mln::math::max(stddev, stddev_color(input_1o16_int_u8)); } - return 0; + return stddev; } int main() { -// unsigned val = stddev_color_16(ANNOTATING_PHOTO_IMG_PATH "/photo01.ppm"); - unsigned val = stddev_color_16(ANNOTATING_LOGO_IMG_PATH "/logo06.ppm"); + typedef boost::filesystem::path t_path; + typedef boost::filesystem::directory_iterator t_iter_path; + + t_path full_path[] = {t_path(ANNOTATING_BILL_IMG_PATH), + t_path(ANNOTATING_FAX_IMG_PATH), + t_path(ANNOTATING_HANDWRITTEN_IMG_PATH), + t_path(ANNOTATING_LOGO_IMG_PATH), + t_path(ANNOTATING_MAP_IMG_PATH), + t_path(ANNOTATING_PHOTO_IMG_PATH), + t_path(ANNOTATING_SCREENSHOT_IMG_PATH), + t_path(ANNOTATING_SLIDE_IMG_PATH), + t_path(ANNOTATING_TYPED_IMG_PATH)}; + + for (int i = 0; i < 9; ++i) + { + std::cerr << "entering " << full_path[i] << std::endl; + std::cout << "entering " << full_path[i] << std::endl; -// std::cout << "nb color : " << val << std::endl; + if (boost::filesystem::exists(full_path[i]) && + boost::filesystem::is_directory(full_path[i])) + { + boost::filesystem::system_complete(full_path[i]); + const t_iter_path end_iter; + float count = 0; + float sum1 = 0; + float sum2 = 0; + + for (t_iter_path dir_iter(full_path[i]); end_iter != dir_iter; ++dir_iter) + { + float val = stddev_color_16(dir_iter->path().string()); + + ++count; + sum1 += val; + sum2 += val*val; + + std::cout << dir_iter->path().string() << " => " << val << std::endl; + } + + float mean = sum1 / count; + float var = ((float)sum2 / count) - (mean * mean); + + std::cout << "mean : " << mean << std::endl; + std::cout << "var : " << var << std::endl; + } + } return 0; } - -- 1.5.6.5

15 years

last-svn-commit-202-gb98887a Fix bugs in regional_maxima routine, work with rgb8 for stats.

by green

* green/demo/labeling/regional_maxima/regional_maxima.cc (compute_stats): New function. * green/demo/labeling/regional_maxima/regional_maxima.cc (demo,main): Update functions to map Theo's view. --- milena/sandbox/ChangeLog | 11 + .../labeling/regional_maxima/regional_maxima.cc | 261 ++++++++++++++++---- 2 files changed, 222 insertions(+), 50 deletions(-) diff --git a/milena/sandbox/ChangeLog b/milena/sandbox/ChangeLog index 7ec7a8a..1ea3546 100644 --- a/milena/sandbox/ChangeLog +++ b/milena/sandbox/ChangeLog @@ -75,6 +75,17 @@ * bin/pgm_to_pbm.cc, * bin/ppm_negate.cc: New. +2009-12-16 Yann Jacquelet <jacquelet(a)lrde.epita.fr> + + Fix bugs in regional_maxima routine, work with rgb8 for stats. + + * green/demo/labeling/regional_maxima/regional_maxima.cc + (compute_stats): New function. + + * green/demo/labeling/regional_maxima/regional_maxima.cc + (demo,main): Update functions to map Theo's view. + + 2009-12-16 Yann Jacquelet <jacquelet(a)lrde.epita.fr> Test the local deviation operator (Millet2008) on the annotating base. diff --git a/milena/sandbox/green/demo/labeling/regional_maxima/regional_maxima.cc b/milena/sandbox/green/demo/labeling/regional_maxima/regional_maxima.cc index 266fbfe..c832486 100644 --- a/milena/sandbox/green/demo/labeling/regional_maxima/regional_maxima.cc +++ b/milena/sandbox/green/demo/labeling/regional_maxima/regional_maxima.cc @@ -2,6 +2,7 @@ #include <iostream> #include <sstream> +#include <boost/format.hpp> #include <mln/img_path.hh> @@ -106,30 +107,31 @@ struct t_labeling_rgbn : mln::Function_v2v< t_labeling_rgbn<n> > } }; +// Version optimisée de mean_values -// version non optimisée de label - -template <unsigned n> -mln::image2d<mln::value::label_8> -label_image(const mln::image2d< mln::value::rgb<n> >& input, - const mln::image3d< mln::value::label_8>& label) +struct t_mean_lbl8_with_rgb8 : mln::Function_v2v< t_mean_lbl8_with_rgb8 > { - mln::image2d<mln::value::label_8> output; + typedef mln::value::rgb8 t_rgb8; + typedef mln::algebra::vec<3,float> t_vec3f; + typedef mln::value::label_8 t_lbl8; + typedef t_lbl8 argument; + typedef t_rgb8 result; + typedef mln::image3d<t_lbl8> t_label; + typedef mln::util::array<t_vec3f> t_mean_array; - initialize(output, input); + const t_mean_array& _mean_array; - mln_piter(mln::image2d< mln::value::label_8 >) po(output.domain()); - mln_piter(mln::image2d< mln::value::rgb<n> >) pi(input.domain()); + t_mean_lbl8_with_rgb8(const t_mean_array& mean_array) : + _mean_array(mean_array) {} - for_all_2(po, pi) + result operator()(const argument& c) const { - const mln::value::rgb<n>& vi = input(pi); + t_rgb8 tmp = _mean_array[c]; - output(po) = mln::opt::at(label, vi.blue(), vi.red(), vi.green()); + return tmp; } +}; - return output; -} template <unsigned n> unsigned unquant(const float& value) @@ -253,6 +255,109 @@ void print_count2(const mln::image2d<mln::value::rgb<n> >& input_rgbn, } } +template <unsigned n> +mln::image2d<mln::value::label_8> +compute_stats(const mln::image2d<mln::value::rgb8>& input_rgb8, + const mln::image2d<mln::value::rgb<n> >& input_rgbn, + const mln::image3d<unsigned>& histo, + const mln::image3d<mln::value::label_8>& label, + const unsigned n_labels) +{ + typedef mln::value::label_8 t_lbl8; + typedef mln::value::rgb<n> t_rgbn; + typedef mln::value::rgb8 t_rgb8; + typedef mln::value::int_u<n> t_int_un; + typedef mln::algebra::vec<3,float> t_vec3f; + typedef mln::accu::math::sum<unsigned,unsigned> t_sum; + typedef mln::accu::stat::mean<t_vec3f,t_vec3f,t_vec3f> t_mean; + typedef mln::image2d<t_lbl8> t_image2d_lbl8; + typedef mln::image2d<t_rgbn> t_image2d_rgbn; + typedef mln::image2d<t_rgb8> t_image2d_rgb8; + typedef mln::image2d<t_int_un> t_image2d_int_un; + typedef mln::util::array<unsigned> t_count_array; + typedef mln::util::array<t_vec3f> t_mean_array; + + mln::util::array<float> abs((unsigned)(n_labels)+1); + mln::util::array<float> rel((unsigned)(n_labels)+1); + unsigned nb = 0; + + for (unsigned i = 0; i <= n_labels; ++i) + { + abs[i] = 0.0; + rel[i] = 0.0; + } + + // COMPUTE THE SUM + t_count_array count = mln::labeling::compute(t_sum(), + histo, + label, + n_labels); + + // COMPUTE THE TOTAL + for (unsigned i = 0; i <= n_labels; ++i) + { + unsigned c = count[i]; + nb += c; + } + + // COMPUTE THE PERCENTAGES + for (unsigned i = 0; i <= n_labels; ++i) + if (0 < count[i]) + { + abs[i] = ((float)count[i] / nb)*100.0; + rel[i] = ((float)count[i] / (nb - count[0]))*100.0; + } + + // COMPUTE THE MEAN + t_image2d_lbl8 label_img = mln::data::transform(input_rgbn, + t_labeling_rgbn<n>(label)); + + t_mean_array mean = mln::labeling::compute(t_mean(), + input_rgb8, + label_img, + n_labels); + + // CORRECT LABEL 0 STATS + rel[0] = 0; + mean[0][0] = 255.0; + mean[0][1] = 255.0; + mean[0][2] = 0.0; + + for (unsigned i = 0; i <= n_labels; ++i) + { + const t_vec3f& mean_v = mean[i]; + + std::cout << boost::format("%2i|" + "r = %6.2f, g = %6.2f, b = %6.2f |" + "c = %7i, %%i = %5.2f, %%c = %5.2f") + % i + % mean_v[0] + % mean_v[1] + % mean_v[2] + % count[i] + % abs[i] + % rel[i] + << std::endl; + /* + std::cout << i << "|(" + << "r = " << unquant<n>(mean_v[0]) << ", " + << "g = " << unquant<n>(mean_v[1]) << ", " + << "b = " << unquant<n>(mean_v[2]) << ")|(" + << "c = " << count[i] << ", " + << "%i= " << abs[i] << "%, " + << "%c= " << rel[i] << "%)" + << std::endl; + */ + } + + t_image2d_rgb8 mean_rgb8 = mln::data::transform(label_img, + t_mean_lbl8_with_rgb8(mean)); + + mln::io::ppm::save(mean_rgb8, "mean.ppm"); + + return label_img; +} + void print_count(const mln::image3d<unsigned>& histo, const mln::image3d<mln::value::label_8>& label, const unsigned n_labels) @@ -320,26 +425,32 @@ void print_count(const mln::image3d<unsigned>& histo, template <unsigned n> struct t_merge_lbl8_with_rgbn : mln::Function_v2v< t_merge_lbl8_with_rgbn<n> > { - typedef mln::value::rgb<n> t_rgbn; - typedef mln::value::label_8 t_lbl8; - typedef t_rgbn argument; - typedef t_rgbn result; - typedef mln::image3d<t_lbl8> t_label; + typedef mln::value::rgb<n> t_rgbn; + typedef mln::value::rgb8 t_rgb8; + typedef mln::value::label_8 t_lbl8; + typedef t_rgb8 argument; + typedef t_rgb8 result; + typedef mln::image3d<t_lbl8> t_label; + typedef mln::fun::v2v::rgb8_to_rgbn<n> t_rgb8_to_rgbn; - const t_label& _label; + const t_label& _label; + const t_rgb8_to_rgbn _convert; t_merge_lbl8_with_rgbn(const t_label& label) : _label(label) {} result operator()(const argument& c) const { - t_rgbn tmp = c; + t_rgb8 tmp_rgb8 = c; + t_rgbn tmp_rgbn = _convert(c); - if (0 == mln::opt::at(_label, c.blue(), c.red(), c.green())) + if (0==mln::opt::at(_label,tmp_rgbn.blue(),tmp_rgbn.red(),tmp_rgbn.green())) { - tmp = mln::literal::black; + // tmp = mln::literal::black; + // tmp = mln::literal::yellow; + tmp_rgb8 = mln::literal::red; } - return tmp; + return tmp_rgb8; } }; @@ -452,13 +563,8 @@ int main2() // n < 8, n is the degree of quantification template <unsigned n> -void demo() +void demo(const std::string& image, const unsigned min_volume) { - const unsigned min_volume = 1000; - //const std::string& image = OLENA_IMG_PATH"/fly.ppm"; - const std::string& image = SCRIBO_PPM_IMG_PATH"/mp00082c_50p.ppm"; - //const std::string& image = OLENA_IMG_PATH"/tiny.ppm"; - typedef mln::value::label_8 t_lbl8; typedef mln::value::int_u8 t_int_u8; typedef mln::value::rgb8 t_rgb8; @@ -473,16 +579,13 @@ void demo() typedef mln::fun::v2v::rgb8_to_rgbn<n> t_rgb8_to_rgbn; typedef mln::accu::meta::stat::histo3d_rgb t_histo3d_fun; - mln::util::timer timer; - // START OF IMAGE PROCESSING CHAIN - timer.start(); - t_image2d_rgb8 input_rgb8; t_image2d_rgbn input_rgbn; - t_image2d_rgbn output_rgbn; - // t_image2d_rgbn mean_rgb5; + t_image2d_rgb8 output_rgb8; + t_image2d_rgb8 mean_rgb8; t_histo3d histo; + t_histo3d histo2; // t_image2d_int_u8 projected; // t_image2d_int_u8 filtered; t_histo3d opened; @@ -490,18 +593,35 @@ void demo() t_image2d_lbl8 label_img; t_image3d_lbl8 dilated; t_lbl8 n_labels; - // t_rgbn value_rgbn; mln::io::ppm::load(input_rgb8, image.c_str()); input_rgbn = mln::data::transform(input_rgb8, t_rgb8_to_rgbn()); histo = mln::data::compute(t_histo3d_fun(), input_rgbn); opened = mln::morpho::opening::volume(histo, mln::c6(), min_volume); label = mln::labeling::regional_maxima(opened, mln::c6(), n_labels); - dilated = mln::morpho::elementary::dilation(label, mln::c26()); + // dilated = mln::morpho::elementary::dilation(label, mln::c26()); - timer.stop(); // END OF IMAGE PROCESSING CHAIN + + // BEGIN DUMPING + label_img = compute_stats(input_rgb8, input_rgbn, histo, label, n_labels); + // mean_rgb8 = mln::labeling::mean_values(input_rgb8, label_img, n_labels); + // mln::io::ppm::save(mean_rgb8, "mean.ppm"); + + output_rgb8 = mln::data::transform(input_rgb8, + t_merge_lbl8_with_rgbn<5>(label)); + + + mln::io::ppm::save(output_rgb8, "merge.ppm"); + mln::io::pgm::save(label_img, "label.pgm"); + + +// histo2 = mln::data::compute(t_histo3d_fun(), mean_rgbn); +// mln::debug::println(histo2); + // END DUMPING + +/* std::ostringstream name; std::ostringstream name2; std::ostringstream name3; @@ -510,17 +630,11 @@ void demo() name2 << "output_rgb" << n << ".ppm"; name3 << "label_img" << n << ".pgm"; - std::cout << "Done in : " << timer.read() << " s" << std::endl; std::cout << "Labels : " << n_labels << std::endl; std::cout << "Name : " << name.str() << std::endl; - mln::io::ppm::save(input_rgbn, name.str()); - - mln::util::timer timer2; +// mln::io::ppm::save(input_rgbn, name.str()); - timer2.start(); - print_count2(input_rgbn, histo, dilated, n_labels); - timer2.stop(); std::cout << "timer2 : " << timer2.read() << std::endl; output_rgbn = mln::data::transform(input_rgbn, t_merge_lbl8_with_rgbn<n>(label)); @@ -531,7 +645,7 @@ void demo() t_labeling_rgbn<n>(label)); // label_img = label_image<n>(input_rgbn, dilated); mln::io::pgm::save(label_img, name3.str()); - +*/ // localiser les couleurs sur l'image (fond en black, le reste) @@ -571,8 +685,8 @@ void demo() */ } - -int main() +/* +int main_old() { demo<2>(); // 2.26 s demo<3>(); // 2.29 s @@ -582,3 +696,50 @@ int main() demo<7>(); // 11.43 s demo<8>(); // 96.19 s } +*/ +void usage() +{ + std::cout << std::endl; + std::cout << "regional_maxima" << std::endl; + std::cout << "image <name>" << std::endl; + std::cout << "quantification {5,6}" << std::endl; + std::cout << "volume 1000" << std::endl; + + // 3 paramètres en entré: + // * l'image + // * la quantification utilisée + // * le volume minimale laissé pour les pics de l'histo + + // sorties: + // * fichier de stats [etiquette|couleur moyenne|% image|% classé] + // * couleur moyenne de l'étiquette + couleur specifique pour étiquette 0 + // * histo sur l'image d'étiquette +} + +int main(int argc, char* args[]) +{ + if (argc != 3) + { + // const std::string& image = OLENA_IMG_PATH"/fly.ppm"; + const std::string& image = SCRIBO_PPM_IMG_PATH"/mp00082c_50p.ppm"; + // const std::string& image = OLENA_IMG_PATH"/tiny.ppm"; + // const std::string image = OLENA_IMG_PATH"/tiny.ppm"; + const unsigned min_volume = 1000; + // const unsigned min_volume = atoi(args[2]); + + switch(args[2][0]) + { + case '2': demo<2>(image, min_volume); break; + case '3': demo<3>(image, min_volume); break; + case '4': demo<4>(image, min_volume); break; + case '5': demo<5>(image, min_volume); break; + case '6': demo<6>(image, min_volume); break; + case '7': demo<7>(image, min_volume); break; + case '8': demo<8>(image, min_volume); break; + default: demo<5>(image, min_volume); break; + // default: usage(); break; + } + } + else + usage(); +} -- 1.5.6.5

15 years

last-svn-commit-203-g7570d06 Fix bug in the direction of projection.

by green

* green/mln/display/display_histo.hh: Fix bug in the direction. * green/mln/display/project_histo.hh: Add spaces to follow the LRDE norm. --- milena/sandbox/ChangeLog | 8 ++++++++ milena/sandbox/green/mln/display/display_histo.hh | 13 +++++++++---- milena/sandbox/green/mln/display/project_histo.hh | 4 ++-- 3 files changed, 19 insertions(+), 6 deletions(-) diff --git a/milena/sandbox/ChangeLog b/milena/sandbox/ChangeLog index 1ea3546..b94de66 100644 --- a/milena/sandbox/ChangeLog +++ b/milena/sandbox/ChangeLog @@ -75,6 +75,14 @@ * bin/pgm_to_pbm.cc, * bin/ppm_negate.cc: New. +2009-12-17 Yann Jacquelet <jacquelet(a)lrde.epita.fr> + + Fix bug in the direction of projection. + + * green/mln/display/display_histo.hh: Fix bug in the direction. + * green/mln/display/project_histo.hh: Add spaces to follow the LRDE + norm. + 2009-12-16 Yann Jacquelet <jacquelet(a)lrde.epita.fr> Fix bugs in regional_maxima routine, work with rgb8 for stats. diff --git a/milena/sandbox/green/mln/display/display_histo.hh b/milena/sandbox/green/mln/display/display_histo.hh index d34914c..1fd5da4 100644 --- a/milena/sandbox/green/mln/display/display_histo.hh +++ b/milena/sandbox/green/mln/display/display_histo.hh @@ -28,10 +28,12 @@ #ifndef MLN_DISPLAY_DISPLAY_HISTO_HH # define MLN_DISPLAY_DISPLAY_HISTO_HH +# include <mln/accu/math/sum.hh> # include <mln/data/stretch.hh> -# include <mln/fun/v2v/log.hh> # include <mln/display/project_histo.hh> -# include <mln/accu/math/sum.hh> +# include <mln/fun/v2v/log.hh> +# include <mln/value/int_u8.hh> + /// \file /// @@ -59,9 +61,12 @@ namespace mln /// /// The 3d histogram is projected in red/green space by /// accumulating around the the blue dimension (green and blue - /// composantes are being correlated). In fact, we sum in along + /// composantes are being correlated). In fact, we sum along /// the blue direction. Then, we stretch value to feet pgm /// format. + /// Direction r = 1 + /// Direction g = 2 + /// Direction b = 0 /// /// \parameter[in] histo the histogram in 3d. /// \result return a equivalent 2d image. @@ -74,7 +79,7 @@ namespace mln typedef value::int_u8 t_int_u8; typedef fun::v2v::log<float> t_log; - image2d<unsigned> proj = project_histo<t_sum,2>(histo); + image2d<unsigned> proj = project_histo<t_sum,0>(histo); image2d<t_int_u8> proj_int = data::stretch(t_int_u8(), data::transform(proj, t_log())); diff --git a/milena/sandbox/green/mln/display/project_histo.hh b/milena/sandbox/green/mln/display/project_histo.hh index 63ea84e..f0e6858 100644 --- a/milena/sandbox/green/mln/display/project_histo.hh +++ b/milena/sandbox/green/mln/display/project_histo.hh @@ -54,7 +54,7 @@ namespace mln image2d<mln_result(A)> project_histo(const image3d<V>& histo); -#ifndef MLN_INCLUDE_ONLY +# ifndef MLN_INCLUDE_ONLY /// \brief Allow the visualization of 3d histogram. /// @@ -86,7 +86,7 @@ namespace mln return accu::image::to_result(histo_accu); } -#endif // ! MLN_INCLUDE_ONLY +# endif // ! MLN_INCLUDE_ONLY } // end of namespace mln::transform -- 1.5.6.5

15 years

last-svn-commit-204-gf3987ed Fix the right behaviour of the regmax software.

by green

* green/demo/labeling/regional_maxima/regional_maxima.cc: Add many modifications to reach the current software state. There is some nomenclatura, some refactoring and debugging. --- milena/sandbox/ChangeLog | 8 + .../labeling/regional_maxima/regional_maxima.cc | 354 ++++++++++++-------- 2 files changed, 214 insertions(+), 148 deletions(-) diff --git a/milena/sandbox/ChangeLog b/milena/sandbox/ChangeLog index b94de66..e1a6235 100644 --- a/milena/sandbox/ChangeLog +++ b/milena/sandbox/ChangeLog @@ -75,6 +75,14 @@ * bin/pgm_to_pbm.cc, * bin/ppm_negate.cc: New. +2009-12-18 Yann Jacquelet<jacquelet(a)lrde.epita.fr> + + Fix the right behaviour of the regmax software. + + * green/demo/labeling/regional_maxima/regional_maxima.cc: Add + many modifications to reach the current software state. There is some + nomenclatura, some refactoring and debugging. + 2009-12-17 Yann Jacquelet <jacquelet(a)lrde.epita.fr> Fix bug in the direction of projection. diff --git a/milena/sandbox/green/demo/labeling/regional_maxima/regional_maxima.cc b/milena/sandbox/green/demo/labeling/regional_maxima/regional_maxima.cc index c832486..65b6608 100644 --- a/milena/sandbox/green/demo/labeling/regional_maxima/regional_maxima.cc +++ b/milena/sandbox/green/demo/labeling/regional_maxima/regional_maxima.cc @@ -27,7 +27,8 @@ // #include <mln/data/stretch.hh> #include <mln/data/transform.hh> -// #include <mln/display/display_histo.hh> +// #include <mln/debug/println.hh> +#include <mln/display/display_histo.hh> #include <mln/fun/v2v/rgb8_to_rgbn.hh> @@ -49,6 +50,7 @@ #include <mln/opt/at.hh> +#include <mln/transform/influence_zone_geodesic.hh> // #include <mln/pw/value.hh> #include <mln/util/array.hh> @@ -255,108 +257,6 @@ void print_count2(const mln::image2d<mln::value::rgb<n> >& input_rgbn, } } -template <unsigned n> -mln::image2d<mln::value::label_8> -compute_stats(const mln::image2d<mln::value::rgb8>& input_rgb8, - const mln::image2d<mln::value::rgb<n> >& input_rgbn, - const mln::image3d<unsigned>& histo, - const mln::image3d<mln::value::label_8>& label, - const unsigned n_labels) -{ - typedef mln::value::label_8 t_lbl8; - typedef mln::value::rgb<n> t_rgbn; - typedef mln::value::rgb8 t_rgb8; - typedef mln::value::int_u<n> t_int_un; - typedef mln::algebra::vec<3,float> t_vec3f; - typedef mln::accu::math::sum<unsigned,unsigned> t_sum; - typedef mln::accu::stat::mean<t_vec3f,t_vec3f,t_vec3f> t_mean; - typedef mln::image2d<t_lbl8> t_image2d_lbl8; - typedef mln::image2d<t_rgbn> t_image2d_rgbn; - typedef mln::image2d<t_rgb8> t_image2d_rgb8; - typedef mln::image2d<t_int_un> t_image2d_int_un; - typedef mln::util::array<unsigned> t_count_array; - typedef mln::util::array<t_vec3f> t_mean_array; - - mln::util::array<float> abs((unsigned)(n_labels)+1); - mln::util::array<float> rel((unsigned)(n_labels)+1); - unsigned nb = 0; - - for (unsigned i = 0; i <= n_labels; ++i) - { - abs[i] = 0.0; - rel[i] = 0.0; - } - - // COMPUTE THE SUM - t_count_array count = mln::labeling::compute(t_sum(), - histo, - label, - n_labels); - - // COMPUTE THE TOTAL - for (unsigned i = 0; i <= n_labels; ++i) - { - unsigned c = count[i]; - nb += c; - } - - // COMPUTE THE PERCENTAGES - for (unsigned i = 0; i <= n_labels; ++i) - if (0 < count[i]) - { - abs[i] = ((float)count[i] / nb)*100.0; - rel[i] = ((float)count[i] / (nb - count[0]))*100.0; - } - - // COMPUTE THE MEAN - t_image2d_lbl8 label_img = mln::data::transform(input_rgbn, - t_labeling_rgbn<n>(label)); - - t_mean_array mean = mln::labeling::compute(t_mean(), - input_rgb8, - label_img, - n_labels); - - // CORRECT LABEL 0 STATS - rel[0] = 0; - mean[0][0] = 255.0; - mean[0][1] = 255.0; - mean[0][2] = 0.0; - - for (unsigned i = 0; i <= n_labels; ++i) - { - const t_vec3f& mean_v = mean[i]; - - std::cout << boost::format("%2i|" - "r = %6.2f, g = %6.2f, b = %6.2f |" - "c = %7i, %%i = %5.2f, %%c = %5.2f") - % i - % mean_v[0] - % mean_v[1] - % mean_v[2] - % count[i] - % abs[i] - % rel[i] - << std::endl; - /* - std::cout << i << "|(" - << "r = " << unquant<n>(mean_v[0]) << ", " - << "g = " << unquant<n>(mean_v[1]) << ", " - << "b = " << unquant<n>(mean_v[2]) << ")|(" - << "c = " << count[i] << ", " - << "%i= " << abs[i] << "%, " - << "%c= " << rel[i] << "%)" - << std::endl; - */ - } - - t_image2d_rgb8 mean_rgb8 = mln::data::transform(label_img, - t_mean_lbl8_with_rgb8(mean)); - - mln::io::ppm::save(mean_rgb8, "mean.ppm"); - - return label_img; -} void print_count(const mln::image3d<unsigned>& histo, const mln::image3d<mln::value::label_8>& label, @@ -561,9 +461,111 @@ int main2() return 0; } +void compute_stats(const mln::image2d<mln::value::rgb8>& i_input_rgb8, + const mln::image2d<mln::value::label_8>& l_input_lbl8, + const mln::image3d<unsigned>& h_histo_rgbn, + const mln::image3d<mln::value::label_8>& l_histo_lbl8, + const mln::value::label_8& n_labels) +{ + typedef mln::algebra::vec<3,float> t_vec3f; + typedef mln::accu::math::sum<unsigned,unsigned> t_sum; + typedef mln::accu::stat::mean<t_vec3f,t_vec3f,t_vec3f> t_mean; + typedef mln::util::array<unsigned> t_count_array; + typedef mln::util::array<t_vec3f> t_mean_array; + + mln::util::array<float> abs((unsigned)(n_labels)+1); + mln::util::array<float> rel((unsigned)(n_labels)+1); + unsigned nb = 0; + + for (unsigned i = 0; i <= n_labels; ++i) + { + abs[i] = 0.0; + rel[i] = 0.0; + } + + // COMPUTE THE SUM + t_count_array count = mln::labeling::compute(t_sum(), + h_histo_rgbn, + l_histo_lbl8, + n_labels); + + // COMPUTE THE TOTAL + for (unsigned i = 0; i <= n_labels; ++i) + { + unsigned c = count[i]; + nb += c; + } + + // COMPUTE THE PERCENTAGES + for (unsigned i = 0; i <= n_labels; ++i) + if (0 < count[i]) + { + abs[i] = ((float)count[i] / nb)*100.0; + rel[i] = ((float)count[i] / (nb - count[0]))*100.0; + } + + // COMPUTE THE MEAN + + t_mean_array mean = mln::labeling::compute(t_mean(), + i_input_rgb8, + l_input_lbl8, + n_labels); + + // CORRECT 0 LABEL STATS + rel[0] = 0; + mean[0][0] = 255.0; + mean[0][1] = 255.0; + mean[0][2] = 0.0; + + // PRINT STATS + for (unsigned i = 0; i <= n_labels; ++i) + { + const t_vec3f& mean_v = mean[i]; + + std::cout << boost::format("%2i|" + "r = %6.2f, g = %6.2f, b = %6.2f |" + "c = %7i, %%i = %5.2f, %%c = %5.2f") + % i + % mean_v[0] + % mean_v[1] + % mean_v[2] + % count[i] + % abs[i] + % rel[i] + << std::endl; + /* + std::cout << i << "|(" + << "r = " << unquant<n>(mean_v[0]) << ", " + << "g = " << unquant<n>(mean_v[1]) << ", " + << "b = " << unquant<n>(mean_v[2]) << ")|(" + << "c = " << count[i] << ", " + << "%i= " << abs[i] << "%, " + << "%c= " << rel[i] << "%)" + << std::endl; + */ + } + + std::cout << std::endl << std::endl; + +} + +void save_histo(const mln::image2d<mln::value::rgb8>& input_rgb8, + const std::string& name) +{ + typedef mln::value::int_u8 t_int_u8; + typedef mln::image2d<t_int_u8> t_image2d_int_u8; + typedef mln::image3d<unsigned> t_histo3d; + typedef mln::accu::meta::stat::histo3d_rgb t_histo3d_fun; + + t_histo3d histo = mln::data::compute(t_histo3d_fun(), input_rgb8); + t_image2d_int_u8 display = mln::display::display_histo3d_unsigned(histo); + mln::io::pgm::save(display, name); +} + + // n < 8, n is the degree of quantification template <unsigned n> -void demo(const std::string& image, const unsigned min_volume) +void demo(const std::string& image, const unsigned min_vol) { typedef mln::value::label_8 t_lbl8; typedef mln::value::int_u8 t_int_u8; @@ -580,45 +582,106 @@ void demo(const std::string& image, const unsigned min_volume) typedef mln::accu::meta::stat::histo3d_rgb t_histo3d_fun; // START OF IMAGE PROCESSING CHAIN - t_image2d_rgb8 input_rgb8; - t_image2d_rgbn input_rgbn; - t_image2d_rgb8 output_rgb8; - t_image2d_rgb8 mean_rgb8; - t_histo3d histo; - t_histo3d histo2; - // t_image2d_int_u8 projected; - // t_image2d_int_u8 filtered; - t_histo3d opened; - t_image3d_lbl8 label; - t_image2d_lbl8 label_img; - t_image3d_lbl8 dilated; - t_lbl8 n_labels; - - mln::io::ppm::load(input_rgb8, image.c_str()); - input_rgbn = mln::data::transform(input_rgb8, t_rgb8_to_rgbn()); - histo = mln::data::compute(t_histo3d_fun(), input_rgbn); - opened = mln::morpho::opening::volume(histo, mln::c6(), min_volume); - label = mln::labeling::regional_maxima(opened, mln::c6(), n_labels); - // dilated = mln::morpho::elementary::dilation(label, mln::c26()); + t_image2d_rgb8 i0_input; // input rgb8 + t_histo3d h0_input; // histo input + + t_image2d_rgbn i1_input; // input rgbn + t_histo3d h1_input; // histo input + + t_histo3d h2_input; // histo opened + t_image3d_lbl8 l2_histo; // label histo + t_image2d_lbl8 l2_input; // label image + t_image2d_rgb8 i2_merge; // merge label + t_image2d_rgb8 i2_mean; // mean label + + t_image3d_lbl8 l3_histo; // label dilated + t_image2d_lbl8 l3_input; // label image + t_image2d_rgb8 i3_merge; // merge label + t_image2d_rgb8 i3_mean; // mean label + + t_image3d_lbl8 l4_histo; // label iz + t_image2d_lbl8 l4_input; // label image + t_image2d_rgb8 i4_merge; // merge label + t_image2d_rgb8 i4_mean; // mean label + + t_image3d_lbl8 l5_histo; // label iz 2 + t_image2d_lbl8 l5_input; // label image + t_image2d_rgb8 i5_merge; // merge label + t_image2d_rgb8 i5_mean; // mean label + + t_lbl8 n_lbl; // nb class + + mln::io::ppm::load(i0_input, image.c_str()); + i1_input = mln::data::transform(i0_input, t_rgb8_to_rgbn()); + h1_input = mln::data::compute(t_histo3d_fun(), i1_input); + h2_input = mln::morpho::opening::volume(h1_input, mln::c6(), min_vol); + l2_histo = mln::labeling::regional_maxima(h2_input, mln::c6(), n_lbl); + l3_histo = mln::morpho::elementary::dilation(l2_histo, mln::c26()); + l4_histo = mln::transform::influence_zone_geodesic(l2_histo, mln::c26()); + l5_histo = mln::transform::influence_zone_geodesic(l2_histo, mln::c26(), 2); // END OF IMAGE PROCESSING CHAIN +// i2_mean = mln::data::transform(l0_input,t_mean_lbl8_with_rgb8(mean_array)); // BEGIN DUMPING - label_img = compute_stats(input_rgb8, input_rgbn, histo, label, n_labels); - // mean_rgb8 = mln::labeling::mean_values(input_rgb8, label_img, n_labels); - // mln::io::ppm::save(mean_rgb8, "mean.ppm"); + save_histo(i0_input, "h0_input.pgm"); + + // without anything + l2_input = mln::data::transform(i1_input, t_labeling_rgbn<n>(l2_histo)); + i2_mean = mln::labeling::mean_values(i0_input, l2_input, n_lbl); + i2_merge = mln::data::transform(i0_input,t_merge_lbl8_with_rgbn<n>(l2_histo)); + + compute_stats(i0_input, l2_input, h1_input, l2_histo, n_lbl); + + mln::io::pgm::save(l2_input, "l2_input.pgm"); + mln::io::ppm::save(i2_mean, "i2_mean.ppm"); + mln::io::ppm::save(i2_merge, "i2_merge.ppm"); + + save_histo(i2_mean, "h2_mean.pgm"); + save_histo(i2_merge, "h2_merge.pgm"); - output_rgb8 = mln::data::transform(input_rgb8, - t_merge_lbl8_with_rgbn<5>(label)); + // with dilation + l3_input = mln::data::transform(i1_input, t_labeling_rgbn<n>(l3_histo)); + i3_mean = mln::labeling::mean_values(i0_input, l3_input, n_lbl); + i3_merge = mln::data::transform(i0_input,t_merge_lbl8_with_rgbn<n>(l3_histo)); + compute_stats(i0_input, l3_input, h1_input, l3_histo, n_lbl); - mln::io::ppm::save(output_rgb8, "merge.ppm"); - mln::io::pgm::save(label_img, "label.pgm"); + mln::io::pgm::save(l3_input, "l3_input.pgm"); + mln::io::ppm::save(i3_mean, "i3_mean.ppm"); + mln::io::ppm::save(i3_merge, "i3_merge.ppm"); + save_histo(i3_mean, "h3_mean.pgm"); + save_histo(i3_merge, "h3_merge.pgm"); -// histo2 = mln::data::compute(t_histo3d_fun(), mean_rgbn); -// mln::debug::println(histo2); + // with geodesic influence + l4_input = mln::data::transform(i1_input, t_labeling_rgbn<n>(l4_histo)); + i4_mean = mln::labeling::mean_values(i0_input, l4_input, n_lbl); + i4_merge = mln::data::transform(i0_input,t_merge_lbl8_with_rgbn<n>(l4_histo)); + + compute_stats(i0_input, l4_input, h1_input, l4_histo, n_lbl); + + mln::io::pgm::save(l4_input, "l4_input.pgm"); + mln::io::ppm::save(i4_mean, "i4_mean.ppm"); + mln::io::ppm::save(i4_merge, "i4_merge.ppm"); + + save_histo(i4_mean, "h4_mean.pgm"); + save_histo(i4_merge, "h4_merge.pgm"); + + // with restricted geodesic influence + l5_input = mln::data::transform(i1_input, t_labeling_rgbn<n>(l5_histo)); + i5_mean = mln::labeling::mean_values(i0_input, l5_input, n_lbl); + i5_merge = mln::data::transform(i0_input,t_merge_lbl8_with_rgbn<n>(l5_histo)); + + compute_stats(i0_input, l5_input, h1_input, l5_histo, n_lbl); + + mln::io::pgm::save(l5_input, "l5_input.pgm"); + mln::io::ppm::save(i5_mean, "i5_mean.ppm"); + mln::io::ppm::save(i5_merge, "i5_merge.ppm"); + + save_histo(i5_mean, "h5_mean.pgm"); + save_histo(i5_merge, "h5_merge.pgm"); // END DUMPING /* @@ -685,25 +748,14 @@ void demo(const std::string& image, const unsigned min_volume) */ } -/* -int main_old() -{ - demo<2>(); // 2.26 s - demo<3>(); // 2.29 s - demo<4>(); // 2.29 s - demo<5>(); // 2.37 s - demo<6>(); // 3.19 s - demo<7>(); // 11.43 s - demo<8>(); // 96.19 s -} -*/ + void usage() { std::cout << std::endl; std::cout << "regional_maxima" << std::endl; std::cout << "image <name>" << std::endl; std::cout << "quantification {5,6}" << std::endl; - std::cout << "volume 1000" << std::endl; + std::cout << "min_volume 1000" << std::endl; // 3 paramètres en entré: // * l'image @@ -712,8 +764,14 @@ void usage() // sorties: // * fichier de stats [etiquette|couleur moyenne|% image|% classé] - // * couleur moyenne de l'étiquette + couleur specifique pour étiquette 0 - // * histo sur l'image d'étiquette + // * image de labels + // * image reconstruite avec la moyenne de chaque classe [i_mean] + // * image reconstruite avec les couleurs de chaque classe [i_merge] + // * histogramme original de l'image d'entrée [h_input] + // * histogramme projeté de mean [h_mean] + // * histogramme projeté de merge [h_merge] + + // * } int main(int argc, char* args[]) -- 1.5.6.5

15 years

last-svn-commit-207-gfe92959 Write the opening volume thresholds for the scribo image mp00082c.ppm.

by green

* green/demo/labeling/regional_maxima/thresholds.txt: New documentation. --- milena/sandbox/ChangeLog | 6 ++++ .../demo/labeling/regional_maxima/thresholds.txt | 27 ++++++++++++++++++++ 2 files changed, 33 insertions(+), 0 deletions(-) create mode 100644 milena/sandbox/green/demo/labeling/regional_maxima/thresholds.txt diff --git a/milena/sandbox/ChangeLog b/milena/sandbox/ChangeLog index 11565c5..48cf813 100644 --- a/milena/sandbox/ChangeLog +++ b/milena/sandbox/ChangeLog @@ -77,6 +77,12 @@ 2009-12-23 Yann Jacquelet <jacquelet(a)lrde.epita.fr> + Write the opening volume thresholds for the scribo image mp00082c.ppm. + + * green/demo/labeling/regional_maxima/thresholds.txt: New documentation. + +2009-12-23 Yann Jacquelet <jacquelet(a)lrde.epita.fr> + Experiment various quantifications on regional maxima labeling. * green/doc/regional_maxima/cmp_quant/h0_input.pgm.gz: New histogram. diff --git a/milena/sandbox/green/demo/labeling/regional_maxima/thresholds.txt b/milena/sandbox/green/demo/labeling/regional_maxima/thresholds.txt new file mode 100644 index 0000000..ddf5ca7 --- /dev/null +++ b/milena/sandbox/green/demo/labeling/regional_maxima/thresholds.txt @@ -0,0 +1,27 @@ +image = 1169 x 1567 = 1831823 + + +% image | min_volume +----------------------- + 0.05 % | 1000.00 + 1.00 % | 18318.23 + 5.00 % | 91591.15 + 10.00 % | 183182.30 + 15.00 % | 274773.45 + 20.00 % | 366364.60 + 25.00 % | 457955.75 + 30.00 % | 549546.90 + 35.00 % | 641138.05 + 40.00 % | 732729.20 + 45.00 % | 824320.35 + 50.00 % | 915911.50 + 55.00 % | 1007502.65 + 60.00 % | 1099093.80 + 65.00 % | 1190684.95 + 70.00 % | 1282276.10 + 75.00 % | 1373867.25 + 80.00 % | 1465458.40 + 85.00 % | 1557049.55 + 90.00 % | 1648640.70 + 95.00 % | 1740231.85 +100.00 % | 1831823.00 -- 1.5.6.5

15 years

last-svn-commit-208-g7c558b5 Split the regional maxima binary in small atomic binaries.

by green

* green/tools/annotating/histo: New directory. * green/tools/annotating/histo/Makefile.am: New Makefile. * green/tools/annotating/histo/histo.cc: New source file. * green/tools/annotating/opening: New directory. * green/tools/annotating/opening/Makefile.am: New Makefile. * green/tools/annotating/opening/opening.cc: New source file. * green/tools/annotating/regmax: New directory. * green/tools/annotating/regmax/Makefile.am: New Makefile. * green/tools/annotating/regmax/regmax.cc: New source file. --- milena/sandbox/ChangeLog | 15 +++ .../annotating/histo}/Makefile.am | 9 +- .../sandbox/green/tools/annotating/histo/histo.cc | 121 ++++++++++++++++++ .../annotating/opening}/Makefile.am | 9 +- .../green/tools/annotating/opening/opening.cc | 79 ++++++++++++ .../annotating/regmax}/Makefile.am | 9 +- .../green/tools/annotating/regmax/regmax.cc | 133 ++++++++++++++++++++ 7 files changed, 360 insertions(+), 15 deletions(-) copy milena/sandbox/green/{exp/annotating/nb_color => tools/annotating/histo}/Makefile.am (96%) create mode 100644 milena/sandbox/green/tools/annotating/histo/histo.cc copy milena/sandbox/green/{exp/annotating/nb_color => tools/annotating/opening}/Makefile.am (96%) create mode 100644 milena/sandbox/green/tools/annotating/opening/opening.cc copy milena/sandbox/green/{exp/annotating/nb_color => tools/annotating/regmax}/Makefile.am (96%) create mode 100644 milena/sandbox/green/tools/annotating/regmax/regmax.cc diff --git a/milena/sandbox/ChangeLog b/milena/sandbox/ChangeLog index 48cf813..c8734e4 100644 --- a/milena/sandbox/ChangeLog +++ b/milena/sandbox/ChangeLog @@ -75,6 +75,21 @@ * bin/pgm_to_pbm.cc, * bin/ppm_negate.cc: New. +2010-01-05 Yann Jacquelet <jacquelet(a)lrde.epita.fr> + + Split the regional maxima binary in small atomic binaries. + + * green/tools/annotating/histo: New directory. + * green/tools/annotating/histo/Makefile.am: New Makefile. + * green/tools/annotating/histo/histo.cc: New source file. + * green/tools/annotating/opening: New directory. + * green/tools/annotating/opening/Makefile.am: New Makefile. + * green/tools/annotating/opening/opening.cc: New source file. + * green/tools/annotating/regmax: New directory. + * green/tools/annotating/regmax/Makefile.am: New Makefile. + * green/tools/annotating/regmax/regmax.cc: New source file. + + 2009-12-23 Yann Jacquelet <jacquelet(a)lrde.epita.fr> Write the opening volume thresholds for the scribo image mp00082c.ppm. diff --git a/milena/sandbox/green/exp/annotating/nb_color/Makefile.am b/milena/sandbox/green/tools/annotating/histo/Makefile.am similarity index 96% copy from milena/sandbox/green/exp/annotating/nb_color/Makefile.am copy to milena/sandbox/green/tools/annotating/histo/Makefile.am index 8e204c6..8cd7511 100644 --- a/milena/sandbox/green/exp/annotating/nb_color/Makefile.am +++ b/milena/sandbox/green/tools/annotating/histo/Makefile.am @@ -6,7 +6,6 @@ # TOOLS # ######### -LOADLIBES= -lboost_filesystem INCLUDES= -I$(HOME)/svn/oln/trunk/milena/sandbox/green #CXXFLAGS= -ggdb -O0 -Wall -W -pedantic -ansi -pipe $(INCLUDES) #CXXFLAGS= -DNDEBUG -O1 -Wall -W -pedantic -ansi -pipe $(INCLUDES) @@ -16,17 +15,17 @@ RM= rm MKDIR= mkdir -p CP= cp -SOURCE_PATTERN= green/exp -BUILD__PATTERN= green/build/exp +SOURCE_PATTERN= green/tools +BUILD__PATTERN= green/build/tools ifeq ($(findstring $(BUILD__PATTERN),$(PWD)), $(BUILD__PATTERN)) # Case where make is done from build directory. SOURCE_DIR= $(subst $(BUILD__PATTERN),$(SOURCE_PATTERN),$(PWD)) -BUILD__DIR= $(PWD)/ +BUILD__DIR= $(PWD) else # Case where make is done from source directory. -SOURCE_DIR= $(PWD)/ +SOURCE_DIR= $(PWD) BUILD__DIR= $(subst $(SOURCE_PATTERN),$(BUILD__PATTERN),$(PWD)) endif diff --git a/milena/sandbox/green/tools/annotating/histo/histo.cc b/milena/sandbox/green/tools/annotating/histo/histo.cc new file mode 100644 index 0000000..ab0b8af --- /dev/null +++ b/milena/sandbox/green/tools/annotating/histo/histo.cc @@ -0,0 +1,121 @@ +// TOOLS ==> Color histogram + +#include <iostream> + +#include <mln/accu/stat/histo3d_rgb.hh> + +#include <mln/core/macros.hh> +#include <mln/core/image/image2d.hh> +#include <mln/core/image/image3d.hh> +#include <mln/core/image/dmorph/image_if.hh> + +#include <mln/data/compute.hh> +#include <mln/data/transform.hh> + +#include <mln/display/display_histo.hh> + +#include <mln/fun/v2v/rgb8_to_rgbn.hh> + +#include <mln/io/dump/save.hh> +#include <mln/io/pbm/load.hh> +#include <mln/io/pbm/save.hh> +#include <mln/io/pgm/load.hh> +#include <mln/io/pgm/save.hh> +#include <mln/io/ppm/load.hh> +#include <mln/io/ppm/save.hh> + +#include <mln/opt/at.hh> + +#include <mln/pw/value.hh> + +#include <mln/value/rgb8.hh> +#include <mln/value/rgb.hh> + + +template <unsigned n> +void mk_histo(const std::string& input, + const std::string& output, + const std::string& histo, + const std::string& mask) +{ + typedef mln::value::int_u8 t_int_u8; + typedef mln::value::rgb8 t_rgb8; + typedef mln::value::rgb<n> t_rgbn; + typedef mln::image2d<t_int_u8> t_image2d_int_u8; + typedef mln::image2d<t_rgb8> t_image2d_rgb8; + typedef mln::image2d<t_rgbn> t_image2d_rgbn; + typedef mln::image2d<bool> t_image2d_bool; + typedef mln::image3d<unsigned> t_histo3d; + typedef mln::fun::v2v::rgb8_to_rgbn<n> t_rgb8_to_rgbn; + typedef mln::accu::meta::stat::histo3d_rgb t_histo3d_fun; + + // START OF IMAGE PROCESSING CHAIN + t_image2d_rgb8 i0_input; // input rgb8 + t_image2d_rgbn i1_input; // input rgbn + t_image2d_bool m0_input; // mask input + t_histo3d h1_input; // histo input + t_image2d_int_u8 p1_histo; // histo proj + + mln::io::ppm::load(i0_input, input.c_str()); + i1_input = mln::data::transform(i0_input, t_rgb8_to_rgbn()); + + if (0 < mask.size()) + { + mln::io::pbm::load(m0_input, mask.c_str()); + h1_input = mln::data::compute(t_histo3d_fun(), + (i1_input | mln::pw::value(m0_input)).rw()); + } + else + { + h1_input = mln::data::compute(t_histo3d_fun(), i1_input); + } + // END OF IMAGE PROCESSING CHAIN + + // BEGIN DUMPING + p1_histo = mln::display::display_histo3d_unsigned(h1_input); + mln::io::dump::save(h1_input, histo.c_str()); + mln::io::pgm::save(p1_histo, output.c_str()); + // END DUMPING +} + + +void usage() +{ + std::cout << std::endl; + std::cout << "histo input.ppm q out.ppm histo.dump [msk.pbm]" << std::endl; + std::cout << "where" << std::endl; + std::cout << "input.ppm is the 8 bits color ppm image" << std::endl; + std::cout << "q is the degree of quanification {2,3,4,5,6,7,8}" << std::endl; + std::cout << "out.pgm is the r/g projection of the histogram" << std::endl; + std::cout << "out.dump is the quantified color histogram" << std::endl; + std::cout << "msk.pbm is the mask which select the pixels" << std::endl; + std::cout << std::endl; +} + +int main(int argc, char* args[]) +{ + if (5 == argc || 6 == argc) + { + const std::string input(args[1]); + const char q = args[2][0]; + const std::string output(args[3]); + const std::string histo(args[4]); + const std::string mask(6 == argc? args[5] : ""); + + switch(q) + { + case '2': mk_histo<2>(input, output, histo, mask); break; + case '3': mk_histo<3>(input, output, histo, mask); break; + case '4': mk_histo<4>(input, output, histo, mask); break; + case '5': mk_histo<5>(input, output, histo, mask); break; + case '6': mk_histo<6>(input, output, histo, mask); break; + case '7': mk_histo<7>(input, output, histo, mask); break; + case '8': mk_histo<8>(input, output, histo, mask); break; + default: usage(); break; + } + } + else + usage(); + + return 0; +} diff --git a/milena/sandbox/green/exp/annotating/nb_color/Makefile.am b/milena/sandbox/green/tools/annotating/opening/Makefile.am similarity index 96% copy from milena/sandbox/green/exp/annotating/nb_color/Makefile.am copy to milena/sandbox/green/tools/annotating/opening/Makefile.am index 8e204c6..8cd7511 100644 --- a/milena/sandbox/green/exp/annotating/nb_color/Makefile.am +++ b/milena/sandbox/green/tools/annotating/opening/Makefile.am @@ -6,7 +6,6 @@ # TOOLS # ######### -LOADLIBES= -lboost_filesystem INCLUDES= -I$(HOME)/svn/oln/trunk/milena/sandbox/green #CXXFLAGS= -ggdb -O0 -Wall -W -pedantic -ansi -pipe $(INCLUDES) #CXXFLAGS= -DNDEBUG -O1 -Wall -W -pedantic -ansi -pipe $(INCLUDES) @@ -16,17 +15,17 @@ RM= rm MKDIR= mkdir -p CP= cp -SOURCE_PATTERN= green/exp -BUILD__PATTERN= green/build/exp +SOURCE_PATTERN= green/tools +BUILD__PATTERN= green/build/tools ifeq ($(findstring $(BUILD__PATTERN),$(PWD)), $(BUILD__PATTERN)) # Case where make is done from build directory. SOURCE_DIR= $(subst $(BUILD__PATTERN),$(SOURCE_PATTERN),$(PWD)) -BUILD__DIR= $(PWD)/ +BUILD__DIR= $(PWD) else # Case where make is done from source directory. -SOURCE_DIR= $(PWD)/ +SOURCE_DIR= $(PWD) BUILD__DIR= $(subst $(SOURCE_PATTERN),$(BUILD__PATTERN),$(PWD)) endif diff --git a/milena/sandbox/green/tools/annotating/opening/opening.cc b/milena/sandbox/green/tools/annotating/opening/opening.cc new file mode 100644 index 0000000..3e1dbf2 --- /dev/null +++ b/milena/sandbox/green/tools/annotating/opening/opening.cc @@ -0,0 +1,79 @@ +// TOOLS ==> histogram filtering + +#include <iostream> + +#include <mln/accu/stat/histo3d_rgb.hh> + +#include <mln/core/macros.hh> +#include <mln/core/alias/neighb3d.hh> +#include <mln/core/image/image2d.hh> +#include <mln/core/image/image3d.hh> + +#include <mln/data/compute.hh> + +#include <mln/display/display_histo.hh> + +#include <mln/io/dump/load.hh> +#include <mln/io/dump/save.hh> +#include <mln/io/pgm/load.hh> +#include <mln/io/pgm/save.hh> + +#include <mln/morpho/opening/volume.hh> + +#include <mln/value/int_u8.hh> + +void mk_opening(const std::string& input, + const unsigned min_vol, + const std::string& output, + const std::string& opened) +{ + typedef mln::value::int_u8 t_int_u8; + typedef mln::image2d<t_int_u8> t_image2d_int_u8; + typedef mln::image3d<unsigned> t_histo3d; + typedef mln::accu::meta::stat::histo3d_rgb t_histo3d_fun; + + // START OF IMAGE PROCESSING CHAIN + t_histo3d h1_input; // histo input + t_histo3d h2_input; // histo input + t_image2d_int_u8 p1_histo; // histo proj + + mln::io::dump::load(h1_input, input.c_str()); + h2_input = mln::morpho::opening::volume(h1_input, mln::c6(), min_vol); + // END OF IMAGE PROCESSING CHAIN + + // BEGIN DUMPING + p1_histo = mln::display::display_histo3d_unsigned(h2_input); + mln::io::dump::save(h2_input, opened.c_str()); + mln::io::pgm::save(p1_histo, output.c_str()); + // END DUMPING +} + + +void usage() +{ + std::cout << std::endl; + std::cout << "opening input.dump v out.dump out.ppm" << std::endl; + std::cout << "where" << std::endl; + std::cout << "input.dump is the 3d color input histo" << std::endl; + std::cout << "v is the minimum size of each composant" << std::endl; + std::cout << "out.pgm is the r/g proj of the opened histogram" << std::endl; + std::cout << "out.dump is the opened histogram" << std::endl; + std::cout << std::endl; +} + +int main(int argc, char* args[]) +{ + if (5 == argc) + { + const std::string input(args[1]); + const unsigned min_vol = atoi(args[2]); + const std::string output(args[3]); + const std::string opened(args[4]); + + mk_opening(input, min_vol, output, opened); + } + else + usage(); + + return 0; +} diff --git a/milena/sandbox/green/exp/annotating/nb_color/Makefile.am b/milena/sandbox/green/tools/annotating/regmax/Makefile.am similarity index 96% copy from milena/sandbox/green/exp/annotating/nb_color/Makefile.am copy to milena/sandbox/green/tools/annotating/regmax/Makefile.am index 8e204c6..8cd7511 100644 --- a/milena/sandbox/green/exp/annotating/nb_color/Makefile.am +++ b/milena/sandbox/green/tools/annotating/regmax/Makefile.am @@ -6,7 +6,6 @@ # TOOLS # ######### -LOADLIBES= -lboost_filesystem INCLUDES= -I$(HOME)/svn/oln/trunk/milena/sandbox/green #CXXFLAGS= -ggdb -O0 -Wall -W -pedantic -ansi -pipe $(INCLUDES) #CXXFLAGS= -DNDEBUG -O1 -Wall -W -pedantic -ansi -pipe $(INCLUDES) @@ -16,17 +15,17 @@ RM= rm MKDIR= mkdir -p CP= cp -SOURCE_PATTERN= green/exp -BUILD__PATTERN= green/build/exp +SOURCE_PATTERN= green/tools +BUILD__PATTERN= green/build/tools ifeq ($(findstring $(BUILD__PATTERN),$(PWD)), $(BUILD__PATTERN)) # Case where make is done from build directory. SOURCE_DIR= $(subst $(BUILD__PATTERN),$(SOURCE_PATTERN),$(PWD)) -BUILD__DIR= $(PWD)/ +BUILD__DIR= $(PWD) else # Case where make is done from source directory. -SOURCE_DIR= $(PWD)/ +SOURCE_DIR= $(PWD) BUILD__DIR= $(subst $(SOURCE_PATTERN),$(BUILD__PATTERN),$(PWD)) endif diff --git a/milena/sandbox/green/tools/annotating/regmax/regmax.cc b/milena/sandbox/green/tools/annotating/regmax/regmax.cc new file mode 100644 index 0000000..2079bc4 --- /dev/null +++ b/milena/sandbox/green/tools/annotating/regmax/regmax.cc @@ -0,0 +1,133 @@ +// TOOLS ==> regmax on histo + +#include <iostream> + +#include <mln/accu/stat/histo3d_rgb.hh> + +#include <mln/core/macros.hh> +#include <mln/core/alias/neighb3d.hh> +#include <mln/core/image/image2d.hh> +#include <mln/core/image/image3d.hh> + +#include <mln/data/compute.hh> + +#include <mln/debug/println.hh> +#include <mln/display/display_histo.hh> + +#include <mln/io/dump/load.hh> +#include <mln/io/dump/save.hh> +#include <mln/io/pgm/load.hh> +#include <mln/io/pgm/save.hh> +#include <mln/io/ppm/save.hh> + +#include <mln/labeling/regional_maxima.hh> + +#include <mln/morpho/opening/volume.hh> + +#include <mln/value/label_8.hh> +#include <mln/value/int_u8.hh> +#include <mln/value/rgb8.hh> + +template <unsigned n> +struct t_labeling_rgbn : mln::Function_v2v< t_labeling_rgbn<n> > +{ + typedef mln::value::rgb<n> t_rgbn; + typedef mln::value::label_8 t_lbl8; + typedef t_rgbn argument; + typedef t_lbl8 result; + typedef mln::image3d<t_lbl8> t_label; + + const t_label& _label; + + t_labeling_rgbn(const t_label& label) : _label(label) {} + + result operator()(const argument& c) const + { + t_lbl8 tmp = mln::opt::at(_label, c.blue(), c.red(), c.green()); + + return tmp; + } +}; + +void mk_regmax(const std::string& input, + const std::string& quant, + const std::string& histo, + const std::string& label, + const std::string& output) +{ + typedef mln::value::label_8 t_lbl8; + typedef mln::value::rgb8 t_rgb8; + typedef mln::value::rgbn t_rgbn; + typedef mln::value::int_u8 t_int_u8; + typedef mln::algebra::vec<3,float> t_v3f; + typedef mln::image2d<t_int_u8> t_image2d_int_u8; + typedef mln::image2d<t_rgb8> t_image2d_rgb8; + typedef mln::image3d<t_lbl8> t_image3d_lbl8; + typedef mln::image2d<t_lbl8> t_image2d_lbl8; + typedef mln::image3d<unsigned> t_histo3d; + typedef mln::accu::meta::stat::histo3d_rgb t_histo3d_fun; + typedef mln::accu::stat::mean<t_v3f,t_v3f,t_v3f> t_mean; + typedef mln::util::array<t_v3f> t_mean_array; + + t_image2d_rgb8 i0_input; // input img + t_image2d_rgbn i1_input; // quant img + t_histo3d h2_input; // opened histo +// t_image2d_int_u8 p2_label; // histo proj + t_image2d_lbl8 p2_label; // histo proj +// t_image2d_rgb8 p2_label; // histo proj + t_image3d_lbl8 l2_histo; // label histo + t_mean_array m2_label; // palette + + t_lbl8 n_lbl; // nb class + + // BEGIN LOADING + mln::io::ppm::load(i0_input, input.c_str()); + mln::io::ppm::load(i1_input, quant.c_str()); + mln::io::dump::load(h2_input, histo.c_str()); + // END LOADING + + // BEGIN IMAGE PROCESSING + l2_histo = mln::labeling::regional_maxima(h2_input, mln::c6(), n_lbl); + // END IMAGE PROCESSING + + // BEGIN SAVING + mln::debug::println(h2_input); + mln::io::dump::save(l2_histo, labeled.c_str()); + + l2_input = mln::data::transform(i1_input, t_labeling_rgbn<n>(l2_histo)); + m2_label = mln::labeling::compute(t_mean(), i0_input, l2_input, n_labels); + p2_label =mln::display::display3_histo3d_unsigned(h2_input,l2_histo,m2_label); + +// mln::io::pgm::save(p2_label, output.c_str()); + mln::io::ppm::save(p2_label, output.c_str()); + std::cout << "Nb classes : " << n_lbl << std::endl; + // END SAVING +} + + +void usage() +{ + std::cout << std::endl; + std::cout << "regmax input.dump out.dump out.ppm" << std::endl; + std::cout << "where" << std::endl; + std::cout << "input.dump is opened histo" << std::endl; + std::cout << "out.pgm is the r/g proj of the opened histogram" << std::endl; + std::cout << "out.dump is the labeled histogram" << std::endl; + std::cout << std::endl; +} + +int main(int argc, char* args[]) +{ + if (4 == argc) + { + const std::string input(args[1]); + const std::string output(args[2]); + const std::string labeled(args[3]); + + mk_regmax(input, output, labeled); + } + else + usage(); + + return 0; +} -- 1.5.6.5

15 years

last-svn-commit-209-g5179913 Build translation table between number of pixels and percentage of pixels in image for the scribo database.

by green

* green/demo/labeling/regional_maxima/threshold.txt: New translation table. --- milena/sandbox/ChangeLog | 8 ++++++++ .../demo/labeling/regional_maxima/thresholds.txt | 15 +++++++++++++++ 2 files changed, 23 insertions(+), 0 deletions(-) diff --git a/milena/sandbox/ChangeLog b/milena/sandbox/ChangeLog index c8734e4..de56c51 100644 --- a/milena/sandbox/ChangeLog +++ b/milena/sandbox/ChangeLog @@ -77,6 +77,14 @@ 2010-01-05 Yann Jacquelet <jacquelet(a)lrde.epita.fr> + Build translation table between number of pixels and percentage of + pixels in image for the scribo database. + + * green/demo/labeling/regional_maxima/threshold.txt: New translation + table. + +2010-01-05 Yann Jacquelet <jacquelet(a)lrde.epita.fr> + Split the regional maxima binary in small atomic binaries. * green/tools/annotating/histo: New directory. diff --git a/milena/sandbox/green/demo/labeling/regional_maxima/thresholds.txt b/milena/sandbox/green/demo/labeling/regional_maxima/thresholds.txt index ddf5ca7..58f3e6a 100644 --- a/milena/sandbox/green/demo/labeling/regional_maxima/thresholds.txt +++ b/milena/sandbox/green/demo/labeling/regional_maxima/thresholds.txt @@ -5,9 +5,24 @@ image = 1169 x 1567 = 1831823 ----------------------- 0.05 % | 1000.00 1.00 % | 18318.23 + 2.00 % | 36636.46 + 3.00 % | 54954.69 + 4.00 % | 73272.92 5.00 % | 91591.15 + 6.00 % | 109909.38 + 7.00 % | 128227.61 + 8.00 % | 146545.84 + 9.00 % | 164864.07 10.00 % | 183182.30 + 11.00 % | 201500.53 + 12.00 % | 219818.76 + 13.00 % | 238136.99 + 14.00 % | 256455.22 15.00 % | 274773.45 + 16.00 % | 293091.68 + 17.00 % | 311409.91 + 18.00 % | 329728.14 + 19.00 % | 348046.37 20.00 % | 366364.60 25.00 % | 457955.75 30.00 % | 549546.90 -- 1.5.6.5

15 years

last-svn-commit-210-g17304c0 Extend the histogram visualization tools for new projection concept.

by green

* green/mln/display/project_histo.hh (project2_histo): New functions that keep the max of the histogram or the class associate to it while projecting along a direction. * green/mln/display/project_histo.hh (project3_histo): New functions that keep the color of the class associate to the histogram maximum while projecting along a direction. * green/mln/display/display_histo.hh: New interface functions for project2_histo and project3_histo. --- milena/sandbox/ChangeLog | 13 + milena/sandbox/green/mln/display/display_histo.hh | 50 +++ milena/sandbox/green/mln/display/project_histo.hh | 344 +++++++++++++++++++++ 3 files changed, 407 insertions(+), 0 deletions(-) diff --git a/milena/sandbox/ChangeLog b/milena/sandbox/ChangeLog index de56c51..f56ea7a 100644 --- a/milena/sandbox/ChangeLog +++ b/milena/sandbox/ChangeLog @@ -77,6 +77,19 @@ 2010-01-05 Yann Jacquelet <jacquelet(a)lrde.epita.fr> + Extend the histogram visualization tools for new projection concept. + + * green/mln/display/project_histo.hh (project2_histo): New functions + that keep the max of the histogram or the class associate to it while + projecting along a direction. + * green/mln/display/project_histo.hh (project3_histo): New functions + that keep the color of the class associate to the histogram maximum + while projecting along a direction. + * green/mln/display/display_histo.hh: New interface functions for + project2_histo and project3_histo. + +2010-01-05 Yann Jacquelet <jacquelet(a)lrde.epita.fr> + Build translation table between number of pixels and percentage of pixels in image for the scribo database. diff --git a/milena/sandbox/green/mln/display/display_histo.hh b/milena/sandbox/green/mln/display/display_histo.hh index 1fd5da4..2ba0b61 100644 --- a/milena/sandbox/green/mln/display/display_histo.hh +++ b/milena/sandbox/green/mln/display/display_histo.hh @@ -33,6 +33,8 @@ # include <mln/display/project_histo.hh> # include <mln/fun/v2v/log.hh> # include <mln/value/int_u8.hh> +# include <mln/value/rgb8.hh> +# include <mln/value/label_8.hh> /// \file @@ -55,6 +57,20 @@ namespace mln image2d<value::int_u8> display_histo3d_unsigned(const image3d<unsigned>& histo); + image2d<value::int_u8> + display2_histo3d_unsigned(const image3d<unsigned>& histo); + + image2d<value::label_8> + display2_histo3d_unsigned(const image3d<unsigned>& histo, + const image3d<value::label_8>& label); + + image2d<value::rgb8> + display3_histo3d_unsigned(const image3d<unsigned>& histo); + + image2d<value::rgb8> + display3_histo3d_unsigned(const image3d<unsigned>& histo, + const image3d<value::label_8>& label); + #ifndef MLN_INCLUDE_ONLY /// \brief Allow the visualization of a 3d histogram by projection. @@ -86,6 +102,40 @@ namespace mln return proj_int; } + image2d<value::int_u8> + display2_histo3d_unsigned(const image3d<unsigned>& histo) + { + image2d<value::int_u8> proj = project2_histo<0>(histo); + + return proj; + } + + image2d<value::label_8> + display2_histo3d_unsigned(const image3d<unsigned>& histo, + const image3d<value::label_8>& label) + { + image2d<value::label_8> proj = project2_histo<0>(histo, label); + + return proj; + } + + image2d<value::rgb8> + display3_histo3d_unsigned(const image3d<unsigned>& histo) + { + image2d<value::rgb8> proj = project3_histo<0>(histo); + + return proj; + } + + image2d<value::rgb8> + display3_histo3d_unsigned(const image3d<unsigned>& histo, + const image3d<value::label_8>& label) + { + image2d<value::rgb8> proj = project3_histo<0>(histo, label); + + return proj; + } + #endif // ! MLN_INCLUDE_ONLY diff --git a/milena/sandbox/green/mln/display/project_histo.hh b/milena/sandbox/green/mln/display/project_histo.hh index f0e6858..d842c70 100644 --- a/milena/sandbox/green/mln/display/project_histo.hh +++ b/milena/sandbox/green/mln/display/project_histo.hh @@ -37,6 +37,12 @@ # include <mln/accu/image/take.hh> # include <mln/accu/image/to_result.hh> +# include <mln/opt/at.hh> + +# include <mln/value/int_u8.hh> +# include <mln/value/rgb8.hh> +# include <mln/value/label_8.hh> + /// \file /// /// \brief Allow the visualization of 3d histogram. @@ -54,6 +60,10 @@ namespace mln image2d<mln_result(A)> project_histo(const image3d<V>& histo); + template <typename A, unsigned direction, typename V> + image2d<mln_result(A)> + project2_histo(const image3d<V>& histo); + # ifndef MLN_INCLUDE_ONLY /// \brief Allow the visualization of 3d histogram. @@ -86,6 +96,340 @@ namespace mln return accu::image::to_result(histo_accu); } + template <unsigned direction> + image2d<value::int_u8> + project2_histo(const image3d<unsigned>& histo) + { + image2d<value::int_u8> result; + + if (0 == direction) // blue + { + image2d<value::int_u8> arg_max(histo.ncols(), histo.nslices()); + + for (unsigned j = 0; j < histo.ncols(); ++j) + for (unsigned i = 0; i < histo.nslices(); ++i) + { + unsigned max = 0; // minimum as possible + signed pos = -1; + + for (unsigned k = 0; k < histo.nrows(); ++k) + { + if (max <= opt::at(histo,i,j,k)) + { + max = opt::at(histo,i,j,k); + pos = k; + } + } + + opt::at(arg_max,i,j) = pos; + } + + result = arg_max; + } + else if (1 == direction) // red + { + image2d<value::int_u8> arg_max(histo.nrows(), histo.nslices()); + + for (unsigned j = 0; j < histo.nslices(); ++j) + for (unsigned i = 0; i < histo.nrows(); ++i) + { + unsigned max = 0; // minimum as possible + signed pos = -1; + + for (unsigned k = 0; k < histo.ncols(); ++k) + { + if (max <= opt::at(histo,i,j,k)) + { + max = opt::at(histo,i,j,k); + pos = k; + } + } + + opt::at(arg_max,i,j) = pos; + } + + result = arg_max; + } + else // 2 == direction // green + { + image2d<value::int_u8> arg_max(histo.nrows(), histo.ncols()); + + for (unsigned j = 0; j < histo.ncols(); ++j) + for (unsigned i = 0; i < histo.nrows(); ++i) + { + unsigned max = 0; // minimum as possible + signed pos = -1; + + for (unsigned k = 0; k < histo.nslices(); ++k) + { + if (max <= opt::at(histo,i,j,k)) + { + max = opt::at(histo,i,j,k); + pos = k; + } + } + + opt::at(arg_max,i,j) = pos; + } + + result = arg_max; + } + + return result; + } + + template <unsigned direction> + image2d<value::label_8> + project2_histo(const image3d<unsigned>& histo, + const image3d<value::label_8>& label) + { + image2d<value::label_8> result; + + if (0 == direction) // blue + { + image2d<value::label_8> arg_max(histo.ncols(), histo.nslices()); + + for (unsigned j = 0; j < histo.ncols(); ++j) + for (unsigned i = 0; i < histo.nslices(); ++i) + { + unsigned max = 0; // minimum as possible + signed pos = -1; + + for (unsigned k = 0; k < histo.nrows(); ++k) + { + if (max <= opt::at(histo,i,j,k)) + { + max = opt::at(histo,i,j,k); + pos = k; + } + } + + opt::at(arg_max,i,j) = opt::at(label,i,j,pos); + } + + result = arg_max; + } + else if (1 == direction) // red + { + image2d<value::label_8> arg_max(histo.nrows(), histo.nslices()); + + for (unsigned j = 0; j < histo.nslices(); ++j) + for (unsigned i = 0; i < histo.nrows(); ++i) + { + unsigned max = 0; // minimum as possible + signed pos = -1; + + for (unsigned k = 0; k < histo.ncols(); ++k) + { + if (max <= opt::at(histo,i,j,k)) + { + max = opt::at(histo,i,j,k); + pos = k; + } + } + + opt::at(arg_max,i,j) = opt::at(label,pos,i,j); + } + + result = arg_max; + } + else // 2 == direction // green + { + image2d<value::label_8> arg_max(histo.nrows(), histo.ncols()); + + for (unsigned j = 0; j < histo.ncols(); ++j) + for (unsigned i = 0; i < histo.nrows(); ++i) + { + unsigned max = 0; // minimum as possible + signed pos = -1; + + for (unsigned k = 0; k < histo.nslices(); ++k) + { + if (max <= opt::at(histo,i,j,k)) + { + max = opt::at(histo,i,j,k); + pos = k; + } + } + + opt::at(arg_max,i,j) = opt::at(label,i,pos,j); + } + + result = arg_max; + } + + return result; + } + + + // FIXME ... determine the color of each class. + template <unsigned direction> + image2d<value::rgb8> + project3_histo(const image3d<unsigned>& histo, + const image3d<value::label_8>& label) + { + image2d<value::rgb8> result; + + if (0 == direction) // blue + { + image2d<value::rgb8> arg_max(histo.ncols(), histo.nslices()); + + for (unsigned j = 0; j < histo.ncols(); ++j) + for (unsigned i = 0; i < histo.nslices(); ++i) + { + unsigned max = 0; // minimum as possible + signed pos = -1; + + for (unsigned k = 0; k < histo.nrows(); ++k) + { + if (max <= opt::at(histo,i,j,k)) + { + max = opt::at(histo,i,j,k); + pos = k; + } + } + + opt::at(arg_max,i,j) = value::rgb8(i,j,pos); + } + + result = arg_max; + } + else if (1 == direction) // red + { + image2d<value::rgb8> arg_max(histo.nrows(), histo.nslices()); + + for (unsigned j = 0; j < histo.nslices(); ++j) + for (unsigned i = 0; i < histo.nrows(); ++i) + { + unsigned max = 0; // minimum as possible + signed pos = -1; + + for (unsigned k = 0; k < histo.ncols(); ++k) + { + if (max <= opt::at(histo,i,j,k)) + { + max = opt::at(histo,i,j,k); + pos = k; + } + } + + opt::at(arg_max,i,j) = value::rgb8(pos,i,j); + } + + result = arg_max; + } + else // 2 == direction // green + { + image2d<value::rgb8> arg_max(histo.nrows(), histo.ncols()); + + for (unsigned j = 0; j < histo.ncols(); ++j) + for (unsigned i = 0; i < histo.nrows(); ++i) + { + unsigned max = 0; // minimum as possible + signed pos = -1; + + for (unsigned k = 0; k < histo.nslices(); ++k) + { + if (max <= opt::at(histo,i,j,k)) + { + max = opt::at(histo,i,j,k); + pos = k; + } + } + + // FIXME ... how to fix the n of rgb + opt::at(arg_max,i,j) = value::rgb8(i,pos,j); + } + + result = arg_max; + } + + return result; + } + + template <unsigned direction> + image2d<value::rgb8> + project3_histo(const image3d<unsigned>& histo) + { + image2d<value::rgb8> result; + + if (0 == direction) // blue + { + image2d<value::rgb8> arg_max(histo.ncols(), histo.nslices()); + + for (unsigned j = 0; j < histo.ncols(); ++j) + for (unsigned i = 0; i < histo.nslices(); ++i) + { + unsigned max = 0; // minimum as possible + signed pos = -1; + + for (unsigned k = 0; k < histo.nrows(); ++k) + { + if (max <= opt::at(histo,i,j,k)) + { + max = opt::at(histo,i,j,k); + pos = k; + } + } + + opt::at(arg_max,i,j) = value::rgb8(i,j,pos); + } + + result = arg_max; + } + else if (1 == direction) // red + { + image2d<value::rgb8> arg_max(histo.nrows(), histo.nslices()); + + for (unsigned j = 0; j < histo.nslices(); ++j) + for (unsigned i = 0; i < histo.nrows(); ++i) + { + unsigned max = 0; // minimum as possible + signed pos = -1; + + for (unsigned k = 0; k < histo.ncols(); ++k) + { + if (max <= opt::at(histo,i,j,k)) + { + max = opt::at(histo,i,j,k); + pos = k; + } + } + + opt::at(arg_max,i,j) = value::rgb8(pos,i,j); + } + + result = arg_max; + } + else // 2 == direction // green + { + image2d<value::rgb8> arg_max(histo.nrows(), histo.ncols()); + + for (unsigned j = 0; j < histo.ncols(); ++j) + for (unsigned i = 0; i < histo.nrows(); ++i) + { + unsigned max = 0; // minimum as possible + signed pos = -1; + + for (unsigned k = 0; k < histo.nslices(); ++k) + { + if (max <= opt::at(histo,i,j,k)) + { + max = opt::at(histo,i,j,k); + pos = k; + } + } + + // FIXME ... how to fix the n of rgb + opt::at(arg_max,i,j) = value::rgb8(i,pos,j); + } + + result = arg_max; + } + + return result; + } + # endif // ! MLN_INCLUDE_ONLY -- 1.5.6.5

15 years

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Olena-patches