added maha dist

cvpr_computedescriptors.m

@@ -22,10 +22,10 @@ OUT_FOLDER = 'descriptors';
 %% and within that folder, create another folder to hold these descriptors
 %% the idea is all your descriptors are in individual folders - within
 %% the folder specified as 'OUT_FOLDER'.
-OUT_SUBFOLDER='avgRGB';
+% OUT_SUBFOLDER='avgRGB';
 % OUT_SUBFOLDER='globalRGBhisto';
 % OUT_SUBFOLDER='spatialColour';
-% OUT_SUBFOLDER='spatialColourTexture';
+OUT_SUBFOLDER='spatialColourTexture';
 
 allfiles=dir (fullfile([DATASET_FOLDER,'/Images/*.bmp']));
 for filenum=1:length(allfiles)
@@ -37,10 +37,10 @@ for filenum=1:length(allfiles)
     fout=[OUT_FOLDER,'/',OUT_SUBFOLDER,'/',fname(1:end-4),'.mat'];%replace .bmp with .mat
     
     %% EXTRACT FUNCTION
-    F=extractAvgRGB(img);
+%     F=extractAvgRGB(img);
 %     F=extractGlobalColHist(img);
 %     F=extractSpatialColour(img);
-%     F=extractSpatialColourTexture(img);
+    F=extractSpatialColourTexture(img);
     save(fout,'F');
     toc
 end

cvpr_visualsearch_pca.m

@@ -27,10 +27,10 @@ DATASET_FOLDER = 'dataset';
 DESCRIPTOR_FOLDER = 'descriptors';
 %% and within that folder, another folder to hold the descriptors
 %% we are interested in working with
-DESCRIPTOR_SUBFOLDER='avgRGB';
+% DESCRIPTOR_SUBFOLDER='avgRGB';
 % DESCRIPTOR_SUBFOLDER='globalRGBhisto';
 % DESCRIPTOR_SUBFOLDER='spatialColour';
-% DESCRIPTOR_SUBFOLDER='spatialColourTexture';
+DESCRIPTOR_SUBFOLDER='spatialColourTexture';
 
 CATEGORIES = ["Farm Animal" 
     "Tree"
@@ -95,23 +95,17 @@ for iterating_category=1:CAT_TOTAL
     %% 2) Select descriptors for category and training data
     category_training_descriptors = [];
     test_descriptors = [];
+    test_categories = [];
     for i=1:NIMG
-        if iterating_category == ALLCATs(i)
+        if (iterating_category == ALLCATs(i)) && (size(category_training_descriptors,1) < MODEL_SIZE)
             category_training_descriptors = [ category_training_descriptors ; ALLFEAT(i,:) ];
         else
             test_descriptors = [ test_descriptors ; ALLFEAT(i,:) ];
+            test_categories = [ test_categories ; ALLCATs(i) ];
         end
     end
     
-    model_descriptors = category_training_descriptors(1:MODEL_SIZE, :);
+    [eig_vct, eig_val, model_mean] = getEigenModel(category_training_descriptors);
-    
-    model_mean = mean(model_descriptors);
-    model_data_min_mean = model_descriptors - repmat(model_mean, MODEL_SIZE, 1);
-
-    C = (model_data_min_mean' * model_data_min_mean) ./ MODEL_SIZE;
-    
-    [eig_vct, eig_val] = eig(C);
-    
     
     TEST_SIZE = size(test_descriptors,1);
     
@@ -119,21 +113,19 @@ for iterating_category=1:CAT_TOTAL
     dst=[];
     for i=1:TEST_SIZE
         candidate=test_descriptors(i,:);
-        query=ALLFEAT(queryimg,:);
+        category=test_categories(i);
-
-        category=ALLCATs(i);
 
         %% COMPARE FUNCTION
-        thedst=compareEuclidean(query, candidate);
+        thedst=compareMahalanobis(eig_vct, eig_val, model_mean, candidate);
         dst=[dst ; [thedst i category]];
     end
     dst=sortrows(dst,1);  % sort the results
 
     %% 4) Calculate PR
-    precision_values=zeros([1, NIMG]);
+    precision_values=zeros([1, TEST_SIZE]);
-    recall_values=zeros([1, NIMG]);
+    recall_values=zeros([1, TEST_SIZE]);
 
-    correct_at_n=zeros([1, NIMG]);
+    correct_at_n=zeros([1, TEST_SIZE]);
 
     query_row = dst(1,:);
     query_category = query_row(1,3);
@@ -141,7 +133,7 @@ for iterating_category=1:CAT_TOTAL
     
     
     %calculate PR for each n
-    for i=1:NIMG
+    for i=1:TEST_SIZE
 
         rows = dst(1:i, :);
 
@@ -182,8 +174,8 @@ for iterating_category=1:CAT_TOTAL
 
 
     %% 5) calculate AP
-    P_rel_n = zeros([1, NIMG]);
+    P_rel_n = zeros([1, TEST_SIZE]);
-    for i = 1:NIMG
+    for i = 1:TEST_SIZE
         precision = precision_values(i);
         i_result_relevant = correct_at_n(i);
 

descriptor/extractSpatialColourTexture.m

@@ -1,7 +1,7 @@
 function F=extractSpatialColourTexture(img)
 
-grid_rows = 10;
+grid_rows = 4;
-grid_columns = 10;
+grid_columns = 4;
 
 img_size = size(img);
 img_rows = img_size(1);
@@ -43,7 +43,7 @@ for i = 1:grid_rows
         avg_vals = extractAvgRGB(img_cell);
         [mag_img, angle_img] = getEdgeInfo(grey_img_cell);
         
-        edge_hist = getEdgeAngleHist(mag_img, angle_img);
+        edge_hist = getEdgeAngleHist(mag_img, angle_img, 6, 0.05);
         
         %concatenate average values into vector
         descriptor = [descriptor edge_hist avg_vals(1) avg_vals(2) avg_vals(3)];

distance/compareMahalanobis.m

@@ -0,0 +1,11 @@
+function dst=compareMahalanobis(vct, val, mean, F2)
+
+x_minus_mean = (F2 - mean)';
+matrices = val' * vct' * x_minus_mean;
+
+x=matrices.^2;
+x=sum(x);
+
+dst=sqrt(sqrt(x));
+
+return;

scratch.m

@@ -1,2 +1,2 @@
 
-getRandomCategoryImage(7)
+loadDescriptors()

util/getEdgeAngleHist.m

@@ -1,7 +1,4 @@
-function F=getEdgeAngleHist(mag_img, angle_img)
+function F=getEdgeAngleHist(mag_img, angle_img, bins, threshold)
-
-bins = 8;
-threshold = 0.05;
 
 dimensions = size(angle_img);
 rows = dimensions(1);
@@ -13,7 +10,7 @@ for i = 1:rows
         if mag_img(i, j) > threshold
             
             bin_value = angle_img(i, j) / (2 * pi);
-            bin_value = bin_value * bins;
+            bin_value = floor(bin_value * bins);
             
             vals = [vals bin_value];
             
@@ -21,5 +18,9 @@ for i = 1:rows
     end
 end
 
-F= histogram(vals, bins, 'Normalization', 'probability').Values;
+if size(vals, 2) == 0
+    F = zeros(1, bins)
+else
+    F= histogram(vals, bins, 'Normalization', 'probability').Values;
+end
 return; 

util/getEigenModel.m

@@ -0,0 +1,12 @@
+function [eig_vct, eig_val, model_mean]=getEigenModel(model_descriptors)
+
+model_size = size(model_descriptors, 1);
+
+model_mean = mean(model_descriptors);
+model_data_min_mean = model_descriptors - repmat(model_mean, model_size, 1);
+
+C = (model_data_min_mean' * model_data_min_mean) ./ model_size;
+
+[eig_vct, eig_val] = eig(C);
+
+return;