Update app.py

app.py

@@ -230,6 +230,7 @@ def get_bbox_from_mask(mask):
 @spaces.GPU
 def get_detection_data(image_data):
     try:
+        # Decode base64 image
         if isinstance(image_data, dict):
             nested_dict = image_data.get("image", {}).get("image", {})
             full_data_url = nested_dict.get("data", "")
@@ -249,10 +250,13 @@ def get_detection_data(image_data):
         img = np.array(img)
         img = cv2.cvtColor(img, cv2.COLOR_RGB2BGR)
 
+        # Run detection
         image = utils.resize(img)
         image_segmentation, objects_data = img_seg.predict(image)
         depthmap, depth_colormap = depth_estimator.make_prediction(image)
 
+        print("✅ Detected Objects:", objects_data)  # Log detected objects
+
         detections = []
         for obj in objects_data:
             cls_id, cls_name, center, mask, _ = obj
@@ -269,6 +273,8 @@ def get_detection_data(image_data):
                 "distance": depth_value
             })
 
+        print("✅ Processed Detections:", detections)  # Log final detections
+
         response = {
             "detections": detections,
             "segmentation_url": save_image_to_url(image_segmentation),
@@ -276,7 +282,7 @@ def get_detection_data(image_data):
             "distance_url": save_image_to_url(utils.draw_depth_info(image, depthmap, objects_data)),
             "point_cloud_url": save_plot_to_url(utils.generate_obj_pcd(depthmap, objects_data)),
             "camera_matrix": get_camera_matrix(depth_estimator),
-            "camera_position": [0, 0, 0]  # Assumed at origin based on camera intrinsics
+            "camera_position": [0, 0, 0]
         }
         return response
 
@@ -285,6 +291,7 @@ def get_detection_data(image_data):
         return {"error": str(e)}
 
 
+
 def cancel():
     CANCEL_PROCESSING = True