Enhance predictions by incorporating center points from neighboring p…

…olygons and considering points outside the current polygon as negative prompt points.

annolid/segmentation/SAM/edge_sam_bg.py

@@ -72,6 +72,24 @@ def random_sample_inside_edges(polygon, num_points):
     return np.array(sampled_points)
 
 
+def random_sample_outside_edges(polygon, num_points):
+    # Randomly sample points inside the edges of the polygon
+    sampled_points = []
+    min_x, min_y, max_x, max_y = polygon.bounds
+
+    for _ in range(num_points):
+        # Generate random point inside the bounding box
+        x = np.random.uniform(min_x, max_x)
+        y = np.random.uniform(min_y, max_y)
+        point = Point(x, y)
+
+        # Check if the point is inside the polygon
+        if not point.within(polygon):
+            sampled_points.append((x, y))
+
+    return np.array(sampled_points)
+
+
 def find_bbox(polygon_points):
     # Convert the list of polygon points to a NumPy array
     points_array = np.array(polygon_points)
@@ -233,14 +251,24 @@ def process_frame(self, frame_number):
             # Randomly sample points inside the edges of the polygon
             points_inside_edges = random_sample_inside_edges(polygon,
                                                              self.num_points_inside_edges)
+            points_outside_edges = random_sample_outside_edges(polygon,
+                                                               self.num_points_inside_edges * 3
+                                                               )
             points_uni = uniform_points_inside_polygon(
                 polygon, self.num_points_inside_edges)
             center_points = self.center_points_dict.get(label,
                                                         MaxSizeQueue(max_size=self.num_center_points))
+
             center_points.enqueue(points[0])
             points = center_points.to_numpy()
             self.center_points_dict[label] = center_points
 
+            # use other instance's center points as negative point prompts
+            other_polygon_center_points = [
+                value for k, v in self.center_points_dict.items() if k != label for value in v]
+            other_polygon_center_points = np.array(
+                [(x[0], x[1]) for x in other_polygon_center_points])
+
             if len(points_inside_edges.shape) > 1:
                 points = np.concatenate(
                     (points, points_inside_edges), axis=0)
@@ -250,6 +278,19 @@ def process_frame(self, frame_number):
                 )
 
             point_labels = [1] * len(points)
+            if len(points_outside_edges) > 1:
+                points = np.concatenate(
+                    (points, points_outside_edges), axis=0
+                )
+                point_labels += [0] * len(points_outside_edges)
+
+            if len(other_polygon_center_points) > 1:
+                points = np.concatenate(
+                    (points, other_polygon_center_points),
+                    axis=0
+                )
+                point_labels += [0] * len(other_polygon_center_points)
+
             polygon = self.edge_sam.predict_polygon_from_points(
                 points, point_labels)