Update app.py

app.py

@@ -27,23 +27,72 @@ pose_model = mp_pose.Pose(static_image_mode=True, model_complexity=2)
 mp_drawing = mp.solutions.drawing_utils
 mp_styles = mp.solutions.drawing_styles
 
-# 🔧 Process function
-def process(image):
+# Define pose connections (edges) for graph structure
+POSE_CONNECTIONS = [
+    # Face
+    (0, 1), (1, 2), (2, 3), (3, 7), (0, 4), (4, 5), (5, 6), (6, 8),
+    # Torso
+    (9, 10), (11, 12), (11, 13), (13, 15), (15, 17), (15, 19), (15, 21),
+    (12, 14), (14, 16), (16, 18), (16, 20), (16, 22), (11, 23), (12, 24),
+    (23, 24),
+    # Left arm
+    (11, 13), (13, 15), (15, 17), (17, 19), (19, 21),
+    # Right arm
+    (12, 14), (14, 16), (16, 18), (18, 20), (20, 22),
+    # Left leg
+    (23, 25), (25, 27), (27, 29), (29, 31), (27, 31),
+    # Right leg
+    (24, 26), (26, 28), (28, 30), (30, 32), (28, 32)
+]
+
+def create_pose_graph_data(pose_landmarks):
+    """Create nodes and edges data structure from pose landmarks"""
+    nodes = {}
+    edges = []
+    
+    # Create nodes
+    for idx, lm in enumerate(pose_landmarks.landmark):
+        name = mp_pose.PoseLandmark(idx).name
+        nodes[idx] = {
+            "id": idx,
+            "name": name,
+            "x": round(lm.x, 4),
+            "y": round(lm.y, 4),
+            "z": round(lm.z, 4),
+            "visibility": round(lm.visibility, 3)
+        }
+    
+    # Create edges based on MediaPipe connections
+    for connection in mp_pose.POSE_CONNECTIONS:
+        start_idx = connection[0]
+        end_idx = connection[1]
+        if start_idx < len(pose_landmarks.landmark) and end_idx < len(pose_landmarks.landmark):
+            edges.append({
+                "from": start_idx,
+                "to": end_idx,
+                "from_name": mp_pose.PoseLandmark(start_idx).name,
+                "to_name": mp_pose.PoseLandmark(end_idx).name
+            })
+    
+    return nodes, edges
+
+def process_pose(image, pose_description=""):
+    """Process pose image and return overlay with pose data"""
     if image is None:
-        return "❌ Please upload an image."
-
+        return None, "❌ Please upload an image.", ""
+    
     # Timestamp-based ID
     ts = datetime.now().strftime("%Y%m%d_%H%M%S")
     pose_id = f"pose_{ts}"
-
+    
     # Convert to OpenCV
     img_bgr = cv2.cvtColor(image, cv2.COLOR_RGB2BGR)
-
+    
     # Detect pose
     results = pose_model.process(cv2.cvtColor(img_bgr, cv2.COLOR_BGR2RGB))
     if not results.pose_landmarks:
-        return "❌ No pose detected."
-
+        return None, "❌ No pose detected.", ""
+    
     # Draw overlay
     overlay = img_bgr.copy()
     mp_drawing.draw_landmarks(
@@ -52,49 +101,151 @@ def process(image):
         mp_pose.POSE_CONNECTIONS,
         landmark_drawing_spec=mp_styles.get_default_pose_landmarks_style()
     )
-
+    
+    # Convert back to RGB for display
+    overlay_rgb = cv2.cvtColor(overlay, cv2.COLOR_BGR2RGB)
+    
     # Save overlay image
     overlay_path = f"pose_images/{pose_id}.png"
     cv2.imwrite(overlay_path, overlay)
-
-    # Extract coordinates
-    pose_coords = {}
-    for idx, lm in enumerate(results.pose_landmarks.landmark):
-        name = mp_pose.PoseLandmark(idx).name
-        pose_coords[name] = {
-            "x": round(lm.x, 4),
-            "y": round(lm.y, 4),
-            "z": round(lm.z, 4),
-            "visibility": round(lm.visibility, 3)
-        }
-
-    # Save to JSON
+    
+    # Create graph data structure
+    nodes, edges = create_pose_graph_data(results.pose_landmarks)
+    
+    # Create pose data summary
+    pose_data = {
+        "pose_id": pose_id,
+        "total_nodes": len(nodes),
+        "total_edges": len(edges),
+        "nodes": nodes,
+        "edges": edges,
+        "description": pose_description if pose_description else "No description provided"
+    }
+    
+    # Save to JSON dataset
     pose_dataset[pose_id] = {
-        "correct pose name": pose_id,
-        "Image with pose overlay": overlay_path,
-        "Pose coordinates": pose_coords,
-        "Pose description": "To be filled"
+        "pose_name": pose_id,
+        "image_path": overlay_path,
+        "pose_description": pose_description,
+        "pose_data": pose_data,
+        "timestamp": ts
     }
-
+    
     with open(json_path, "w") as f:
         json.dump(pose_dataset, f, indent=2)
-
-    # Return basic preview
-    preview = f"✅ Pose saved as `{pose_id}` with {len(pose_coords)} joints.\n"
-    for joint, v in list(pose_coords.items())[:5]:  # show first 5
-        preview += f"{joint}: x={v['x']} y={v['y']} z={v['z']}\n"
-
-    preview += f"\n🖼 Overlay image saved at `{overlay_path}`"
-
-    return preview
-
-# ✅ Interface-style Gradio UI
-interface = gr.Interface(
-    fn=process,
-    inputs=gr.Image(type="numpy", label="Upload Pose Image"),
-    outputs="text",
-    title="🧘 Pose Analysis with MediaPipe",
-    description="Upload a yoga or archery pose image. This tool will extract pose keypoints using MediaPipe and save them in a JSON file."
-)
-
-interface.launch()
+    
+    # Format pose data for display
+    data_display = f"""
+🎯 **Pose Analysis Results**
+
+📊 **Graph Structure:**
+- Total Nodes (Keypoints): {len(nodes)}
+- Total Edges (Connections): {len(edges)}
+
+📝 **Pose Description:** {pose_description if pose_description else "No description provided"}
+
+🔍 **Key Nodes (First 10):**
+"""
+    
+    for i, (idx, node) in enumerate(list(nodes.items())[:10]):
+        data_display += f"• {node['name']}: ({node['x']:.3f}, {node['y']:.3f}, {node['z']:.3f}) [visibility: {node['visibility']}]\n"
+    
+    if len(nodes) > 10:
+        data_display += f"... and {len(nodes) - 10} more nodes\n"
+    
+    data_display += f"""
+🔗 **Sample Connections:**
+"""
+    
+    for i, edge in enumerate(edges[:5]):
+        data_display += f"• {edge['from_name']} → {edge['to_name']}\n"
+    
+    if len(edges) > 5:
+        data_display += f"... and {len(edges) - 5} more connections\n"
+    
+    data_display += f"""
+💾 **Data Saved:**
+- Image: {overlay_path}
+- JSON: {json_path}
+- Pose ID: {pose_id}
+"""
+    
+    return overlay_rgb, data_display, f"✅ Pose '{pose_id}' saved successfully!"
+
+def save_with_description(image, description):
+    """Save pose with description"""
+    if image is None:
+        return None, "❌ Please upload an image first.", "❌ No image to process"
+    
+    return process_pose(image, description)
+
+# Create Gradio interface
+with gr.Blocks(title="🧘 Advanced Pose Analysis Tool") as demo:
+    gr.Markdown("# 🧘 Advanced Pose Analysis with MediaPipe")
+    gr.Markdown("Upload a yoga, archery, or any pose image to extract keypoints and analyze the pose structure as nodes and edges.")
+    
+    with gr.Row():
+        with gr.Column(scale=1):
+            # Input section
+            gr.Markdown("## 📤 Input")
+            input_image = gr.Image(type="numpy", label="Upload Pose Image")
+            pose_description = gr.Textbox(
+                label="Pose Description",
+                placeholder="Enter a description of the pose (e.g., 'Warrior II pose with arms extended')",
+                lines=3
+            )
+            
+            with gr.Row():
+                analyze_btn = gr.Button("🔍 Analyze Pose", variant="primary")
+                save_btn = gr.Button("💾 Save with Description", variant="secondary")
+        
+        with gr.Column(scale=1):
+            # Output section
+            gr.Markdown("## 📊 Results")
+            output_image = gr.Image(label="Pose with MediaPipe Overlay")
+            status_text = gr.Textbox(label="Status", lines=1)
+    
+    # Pose data display
+    gr.Markdown("## 📋 Pose Data (Nodes & Edges)")
+    pose_data_display = gr.Textbox(
+        label="Pose Analysis Data",
+        lines=15,
+        max_lines=20,
+        show_copy_button=True
+    )
+    
+    # Button actions
+    analyze_btn.click(
+        fn=lambda img: process_pose(img, ""),
+        inputs=[input_image],
+        outputs=[output_image, pose_data_display, status_text]
+    )
+    
+    save_btn.click(
+        fn=save_with_description,
+        inputs=[input_image, pose_description],
+        outputs=[output_image, pose_data_display, status_text]
+    )
+    
+    # Auto-analyze on image upload
+    input_image.change(
+        fn=lambda img: process_pose(img, ""),
+        inputs=[input_image],
+        outputs=[output_image, pose_data_display, status_text]
+    )
+    
+    gr.Markdown("""
+    ## 📖 How to Use:
+    1. **Upload Image**: Upload a pose image using the image uploader
+    2. **Auto Analysis**: The pose will be automatically analyzed showing keypoints
+    3. **Add Description**: Enter a description of the pose in the text box
+    4. **Save**: Click "Save with Description" to save the pose data with your description
+    
+    ## 📊 Data Structure:
+    - **Nodes**: 33 body keypoints (face, torso, arms, legs) with 3D coordinates
+    - **Edges**: Connections between keypoints following human body structure
+    - **Visibility**: Confidence score for each keypoint detection
+    """)
+
+# Launch the interface
+demo.launch(share=True)