Adicionando opcao Ground Truth

src/interface/tk/tk_menu.py

@@ -38,7 +38,7 @@ class Menu(object):
         self._items_menu = []
         
 
-    def add_command(self, label, action, shortcut):
+    def add_command(self, label, action, shortcut = None):
         """Add a new option to menu.
 
         Parameters
@@ -66,7 +66,7 @@ class Menu(object):
         """
         self._items_menu.append( Separator(self) )
         
-    def add_check_button(self, label, action, shortcut):
+    def add_check_button(self, label, action, shortcut = None, default_state = True):
         """Add a new check button to menu.
 
         Parameters
@@ -77,8 +77,10 @@ class Menu(object):
             Callback to be executed on check button click.
         shortcut : string, optional, default = None
             Check button shortcut number ou letter. Not case sensitive.
+        default_state : boolean, optional, default = True
+            Initial state of check button. If true set to on.
         """
-        self._items_menu.append( CheckButton(self, label, action, shortcut) )
+        self._items_menu.append( CheckButton(self, label, action, shortcut, default_state) )
 
     def render(self, menubar):
         """Render the menu in GUI.

src/interface/tk_interface.py

@@ -87,7 +87,7 @@ class TkInterface(Interface):
         """
         self._menus[-1].add_separator( )
         
-    def add_check_button(self, label, action, shortcut = None):
+    def add_check_button(self, label, action, shortcut = None, default_state = True):
         """Add a check button to last menu added.
 
         Parameters
@@ -96,10 +96,12 @@ class TkInterface(Interface):
             Check button option label.
         action : function
             Callback to be executed on check button click.
-        shortcut : string
+        shortcut : string, optional, default = None
             Check button shortcut number ou letter. Not case sensitive.
+        default_state : boolean, optional, default = True
+            Initial state of check button. If true set to on.
         """
-        self._menus[-1].add_check_button(label = label, action = lambda *_: self._apply( action ), shortcut = shortcut)
+        self._menus[-1].add_check_button(label = label, action = lambda *_: self._apply( action ), shortcut = shortcut, default_state = default_state)
 
     def render_menu(self):
         """Add to window GUI the last menu added.

src/main.py

@@ -6,7 +6,7 @@
     
     This file is the main code of Pynovisao.
     To add a new action ( functionality ) to this software, you must implement the method body of action in file pynovisao.py 
-    and bind the method to menu, using the call tk.add_command.
+    and bind the method to menu, using the call tk.add_command or tk.add_check_button.
     
     To add a new Feature Extractor, Segmenter or Classifier, place it in correct directory and extend its abstract class.
     
@@ -63,7 +63,7 @@ if __name__ == "__main__":
     tk.add_command("Add new class", act.add_class, 'A')
     tk.add_command("Set dataset path", act.set_dataset_path, 'd')
     tk.add_separator()
-    tk.add_check_button("Dataset generator", act.toggle_dataset_generator, 't')
+    tk.add_check_button("Dataset generator", act.toggle_dataset_generator)
     
     tk.add_menu("Segmentation")
     tk.add_command("Choose segmenter", act.select_segmenter)
@@ -80,6 +80,8 @@ if __name__ == "__main__":
     tk.add_command("Choose classifier", act.select_classifier)
     tk.add_command("Configure", act.configure_classifier)
     tk.add_separator()
+    tk.add_check_button("Ground Truth", act.toggle_ground_truth, default_state = False)
+    tk.add_separator()
     tk.add_command("Cross Validation", act.cross_validation, 'X')
     tk.add_command("Experimenter All", act.experimenter_all, 'p')
     tk.add_separator()

src/pynovisao.py

@@ -9,6 +9,7 @@
 """
 
 from collections import OrderedDict
+import numpy as np
 
 import interface
 from interface.interface import InterfaceException as IException
@@ -17,6 +18,7 @@ import segmentation
 import extraction
 from extraction import FeatureExtractor
 import classification
+from classification import Classifier
 
 import util
 from util.config import Config
@@ -58,6 +60,8 @@ class Act(object):
         self._init_classes(args["classes"], args["colors"])
         
         self._dataset_generator = True
+        self._ground_truth = False
+        self._gt_segments = None
 
     
     def _init_dataset(self, directory):
@@ -125,6 +129,10 @@ class Act(object):
                     self.tk.append_log("Painting segment: %0.3f seconds", run_time)
                     self.tk.refresh_image(self._image)
                     
+                    if self._ground_truth == True:
+                        self._gt_segments[idx_segment] = self.classes[self._current_class]["name"].value
+
+                    elif self._dataset_generator == True:
                         filepath = f.save_class_image(segment, self.dataset, self.classes[self._current_class]["name"].value, self._image_name, idx_segment)
                         if filepath:
                             self.tk.append_log("\nSegment saved in %s", filepath)
@@ -141,6 +149,7 @@ class Act(object):
             self._const_image = self._image
             
             self.segmenter.reset()
+            self._gt_segments = None
 
         
     def restore_image(self):
@@ -151,6 +160,7 @@ class Act(object):
             self.tk.refresh_image(self._const_image)
             
             self.segmenter.reset()
+            self._gt_segments = None
         
     def close_image(self):
         """Close the image.
@@ -354,6 +364,8 @@ class Act(object):
         self._image, run_time = self.segmenter.run(self._const_image)
         self.tk.append_log("Time elapsed: %0.3f seconds", run_time)
         
+        self._gt_segments = [None]*(max(self.segmenter.get_list_segments())+1)
+        
         self.tk.refresh_image(self._image)
 
 
@@ -488,6 +500,7 @@ class Act(object):
             self._image, _ = self.segmenter.run(self._const_image)
             self.tk.refresh_image(self._image)        
             list_segments = self.segmenter.get_list_segments()
+            self._gt_segments = [None]*(max(list_segments)+1)
         
         # Train the classifier ( this program does not perform the training of ConvNets ).
         if self.classifier.must_train():
@@ -530,6 +543,9 @@ class Act(object):
         
         self.tk.append_log("Painting segments... (%0.3f seconds)", (TimeUtils.get_time() - start_time))
         
+        # If ground truth mode, show alternative results
+        if self._ground_truth == True:
+            return self._show_ground_truth(list_segments, len_segments, labels, start_time)
 
         # Create a popup with results of classification.
         popup_info = "%s\n" % str(self.classifier.get_summary_config())
@@ -556,6 +572,58 @@ class Act(object):
         self.tk.append_log("Time elapsed: %0.3f seconds", (end_time - start_time))
         
     
+    def _show_ground_truth(self, list_segments, len_segments, labels, start_time):
+        """Paint only wrong classified segments and show ground truth confusion matrix.
+        
+        Parameters
+        ----------
+        list_segments : list of integer
+            List of index segments.
+        len_segments : list of integer
+            List of segments sizes.
+        labels : list of string
+            List of predicted class name for each segment.
+        start_time : floating point
+            Start time of classification.
+        """
+        classes = list(set(labels))
+        classes.sort()
+        
+        n_segments = len(labels)
+        spx_matrix = np.zeros((len(classes), len(classes)), np.int) 
+        px_matrix = np.zeros((len(classes), len(classes)), np.int) 
+
+        # Create the confusion matrix and paint wrong classified segments individually.
+        for idx_segment in list_segments:
+            if self._gt_segments[idx_segment] is not None:
+                gt_class = classes.index(self._gt_segments[idx_segment])
+                predicted_class = classes.index(labels[idx_segment])
+                
+                spx_matrix[ gt_class ][ predicted_class ] += 1
+                px_matrix[ gt_class ][ predicted_class ] += len_segments[idx_segment]
+        
+                if gt_class != predicted_class:
+                    self._image, _ = self.segmenter.paint_segment(self._image, self.get_class_by_name(labels[idx_segment])["color"].value, idx_segment=[idx_segment], border=False)
+        
+        # Create a popup with results of classification.
+        popup_info = "%s\n" % str(self.classifier.get_summary_config())
+        popup_info += Classifier.confusion_matrix(classes, spx_matrix, "Superpixels")
+        popup_info += Classifier.confusion_matrix(classes, px_matrix, "PixelSum")
+        
+        self.tk.refresh_image(self._image)
+        self.tk.popup(popup_info)
+
+        end_time = TimeUtils.get_time()
+            
+        self.tk.append_log("\nClassification finished")
+        self.tk.append_log("Time elapsed: %0.3f seconds", (end_time - start_time))
+        
+
+    def toggle_ground_truth(self):
+        """Enable/disable ground truth mode.
+        """
+        self._ground_truth = not self._ground_truth
+        
     def cross_validation(self):
         """Run a cross validation on all generated segments in image dataset.
         

src/segmentation/skimage_segmenter.py

@@ -157,7 +157,7 @@ class SkimageSegmenter(object):
             idx_segment = [self._segments[py, px]]
         height, width, channels = self._original_image.shape
         
-        # Create a mask, painting black all pixels outside of segment and white the pixels inside
+        # Create a mask, painting black all pixels outside of segments and white the pixels inside
         mask_segment = np.zeros(self._original_image.shape[:2], dtype="uint8")
         for idx in idx_segment:
             mask_segment[self._segments == idx] = 255
@@ -172,7 +172,7 @@ class SkimageSegmenter(object):
             colored_image = cv2.addWeighted(image, 0.7, class_color, 0.3, 0)
         colored_image = cv2.bitwise_and(colored_image, colored_image, mask=mask_segment)
         
-        # Create a new image keeping the painting only in pixels inside of segment
+        # Create a new image keeping the painting only in pixels inside of segments
         new_image = image if clear == False else self._original_image
         new_image = cv2.bitwise_and(new_image, new_image, mask=mask_inv)
         mask_segment[:] = 255