def hot_points_get(self, image_array):
          
          image_data = [0]*self.len_array
          
          for i in range(self.len_array):
               image_data[i] = ((image_array[i][0] + image_array[i][1] + image_array[i][2]) - variables.sensible_3)
          
          image_scanned = numpy.array(image_data)
          image_scanned = numpy.reshape(image_scanned, (self.im_height, self.im_width))
          
          #We clean the old array
          self.hot_points = []
          
          image_scanned = self.border_make(image_scanned)
               
          #We follow the matrice
          for ki in range(self.hei):
               for kj in range(self.wid):
                    
                    i = (ki * self.minSize) + 2
                    j = (kj * self.minSize) + 2
                    
               
                    #There is a hot point !
                    if image_scanned[i][j] > 0:
                         #Init a new zone
                         self.Cadre = [j, i, j, i] #we set Min_x, Min_y, Max_x, Max_y
                         self.Barycentre = [j, i] #x and y barycentre
                         self.Poids = image_scanned[i][j]  #Weight of actual barycentre
                         
                         image_scanned[i][j] = 0 #Kill this point
                         
                         #Get the hot zone, (by reccurence)
                         self.hot_point_zone(j, i, image_scanned)
                         
                         #Get the size of the hot point, and add him if he is not too small
                         Cadre_Size = (self.Cadre[2] - self.Cadre[0]) * (self.Cadre[3] - self.Cadre[1])
                         
                         if Cadre_Size > self.minSize:
                              Temp_Info = [self.Cadre, self.Barycentre, self.Poids]
                              self.hot_points.append(Temp_Info)
                         
          return self.hot_points