Update app.py

app.py

@@ -49,6 +49,37 @@ def draw_bounding_boxes(frame, detections):
     frame_with_boxes = cv2.cvtColor(np.array(pil_image), cv2.COLOR_RGB2BGR)
     return frame_with_boxes
 
+def create_output_writer(cap, output_path):
+    """
+    Create video writer with different codecs, trying multiple options
+    """
+    frame_width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
+    frame_height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
+    fps = int(cap.get(cv2.CAP_PROP_FPS))
+    
+    # Try different codecs
+    codecs = [
+        ('mp4v', '.mp4'),
+        ('avc1', '.mp4'),
+        ('XVID', '.avi'),
+        ('MJPG', '.avi')
+    ]
+    
+    for codec, ext in codecs:
+        try:
+            output_file = os.path.splitext(output_path)[0] + ext
+            fourcc = cv2.VideoWriter_fourcc(*codec)
+            out = cv2.VideoWriter(output_file, fourcc, fps, (frame_width, frame_height))
+            
+            if out is not None and out.isOpened():
+                return out, output_file
+                
+        except Exception as e:
+            print(f"Failed with codec {codec}: {str(e)}")
+            continue
+            
+    raise ValueError("Could not initialize any video codec")
+
 def process_video(video_path, progress=gr.Progress()):
     """
     Process the video file and return the path to the processed video
@@ -59,21 +90,17 @@ def process_video(video_path, progress=gr.Progress()):
         if not cap.isOpened():
             raise ValueError("Could not open video file")
 
-        # Get video properties
-        fps = int(cap.get(cv2.CAP_PROP_FPS))
-        frame_width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
-        frame_height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
         total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
 
-        # Create output video file
-        output_path = os.path.join(tempfile.gettempdir(), 'output_video.mp4')
+        # Create output directory if it doesn't exist
+        output_dir = os.path.join(os.path.expanduser("~"), "Videos", "ObjectDetection")
+        os.makedirs(output_dir, exist_ok=True)
         
-        # Initialize video writer with H264 codec
-        fourcc = cv2.VideoWriter_fourcc(*'avc1')
-        out = cv2.VideoWriter(output_path, fourcc, fps, (frame_width, frame_height))
-
-        if not out.isOpened():
-            raise ValueError("Could not create output video file")
+        # Create output path
+        output_path = os.path.join(output_dir, "output_video.mp4")
+        
+        # Initialize video writer
+        out, output_path = create_output_writer(cap, output_path)
 
         frame_count = 0
         process_every_n_frames = 1  # Process every frame
@@ -137,10 +164,10 @@ def detect_objects_in_video(video):
 demo = gr.Interface(
     fn=detect_objects_in_video,
     inputs=[
-        gr.Video(label="Upload Video", format="mp4")
+        gr.Video(label="Upload Video")
     ],
     outputs=[
-        gr.Video(label="Processed Video", format="mp4")
+        gr.Video(label="Processed Video")
     ],
     title="@GenAILearniverse Project: Video Object Detection",
     description="""