webcam

app.py

@@ -193,6 +193,37 @@ def format_results(result, confidence_threshold: float) -> str:
     return results_text
 
 
+def process_webcam_stream(frame, model_name: str, confidence_threshold: float):
+    """
+    Process webcam stream frame by frame.
+    
+    Args:
+        frame: Camera frame as numpy array
+        model_name: Name of the model to use
+        confidence_threshold: Confidence threshold for predictions
+        
+    Returns:
+        Processed frame as PIL Image
+    """
+    if frame is None:
+        return None
+    
+    try:
+        # Load model
+        model = load_model(model_name)
+        
+        # Run inference
+        result = model(frame)
+        
+        # Visualize results
+        result_image = visualizer.render(frame, result)
+        
+        return result_image
+    except Exception as e:
+        print(f"[ERROR] Webcam stream error: {e}")
+        return frame
+
+
 def create_gradio_interface():
     """
     Create and configure the Gradio interface.
@@ -208,53 +239,79 @@ def create_gradio_interface():
     
     with gr.Blocks(title="Object Detection with model_api") as demo:
         gr.Markdown("# 🎯 Object Detection with model_api")
-        gr.Markdown("Upload an image and select a model to perform object detection using OpenVINO and model_api")
+        gr.Markdown("Upload an image, use your webcam, or select a model to perform object detection using OpenVINO and model_api")
         
+        # Model selection (shared across tabs)
         with gr.Row():
-            with gr.Column(scale=1):
-                input_image = gr.Image(
-                    label="Input Image",
-                    type="numpy",
-                    height=400
-                )
-                
-                model_dropdown = gr.Dropdown(
-                    choices=available_models,
-                    value=available_models[0] if available_models else None,
-                    label="Select Model",
-                    info="Choose a model from the models/ folder"
-                )
-                
-                confidence_slider = gr.Slider(
-                    minimum=0.0,
-                    maximum=1.0,
-                    value=0.3,
-                    step=0.05,
-                    label="Confidence Threshold",
-                    info="Minimum confidence for displaying predictions"
-                )
-                
-                classify_btn = gr.Button("🚀 Run Inference", variant="primary")
+            model_dropdown = gr.Dropdown(
+                choices=available_models,
+                value=available_models[0] if available_models else None,
+                label="Select Model",
+                info="Choose a model from the models/ folder"
+            )
+            
+            confidence_slider = gr.Slider(
+                minimum=0.0,
+                maximum=1.0,
+                value=0.3,
+                step=0.05,
+                label="Confidence Threshold",
+                info="Minimum confidence for displaying predictions"
+            )
+        
+        # Create tabs for different input sources
+        with gr.Tabs():
+            # Image Tab
+            with gr.TabItem("📷 Image Upload"):
+                with gr.Row():
+                    with gr.Column(scale=1):
+                        input_image = gr.Image(
+                            label="Input Image",
+                            type="numpy",
+                            height=400
+                        )
+                        
+                        classify_btn = gr.Button("🚀 Run Inference", variant="primary")
+                    
+                    with gr.Column(scale=1):
+                        output_image = gr.Image(
+                            label="Detection Result",
+                            type="pil",
+                            show_label=False,
+                            height=400
+                        )
+                        
+                        detections_output = gr.Textbox(
+                            label="Detected Objects",
+                            lines=8,
+                            max_lines=15
+                        )
+                        
+                        metrics_output = gr.Textbox(
+                            label="Performance Metrics",
+                            lines=8,
+                            max_lines=15
+                        )
             
-            with gr.Column(scale=1):
-                output_image = gr.Image(
-                    label="Detection Result",
-                    type="pil",
-                    show_label=False,
-                    height=400
-                )
-                
-                detections_output = gr.Textbox(
-                    label="Detected Objects",
-                    lines=8,
-                    max_lines=15
-                )
-                
-                metrics_output = gr.Textbox(
-                    label="Performance Metrics",
-                    lines=8,
-                    max_lines=15
-                )
+            # Webcam Tab
+            with gr.TabItem("🎥 Webcam"):
+                with gr.Row():
+                    with gr.Column(scale=1):
+                        webcam_input = gr.Image(
+                            sources=["webcam"],
+                            label="Webcam Feed",
+                            type="numpy",
+                            streaming=True,
+                            height=400
+                        )
+                    
+                    with gr.Column(scale=1):
+                        webcam_output = gr.Image(
+                            label="Live Detection",
+                            type="numpy",
+                            show_label=False,
+                            height=400
+                        )
         
         # Examples section
         gr.Markdown("## 📸 Examples")
@@ -271,12 +328,21 @@ def create_gradio_interface():
             cache_examples=False
         )
         
-        # Connect the button to the inference function
+        # Connect the button to the inference function for image upload
         classify_btn.click(
             fn=run_inference,
             inputs=[input_image, model_dropdown, confidence_slider],
             outputs=[output_image, detections_output, metrics_output]
         )
+        
+        # Connect webcam stream processing
+        webcam_input.stream(
+            fn=process_webcam_stream,
+            inputs=[webcam_input, model_dropdown, confidence_slider],
+            outputs=webcam_output,
+            stream_every=0.1,
+            time_limit=300
+        )
     
     return demo