Update app.py

app.py

@@ -1,3 +1,4 @@
+# app.py
 import gradio as gr
 import numpy as np
 import cv2
@@ -12,9 +13,6 @@ import io
 
 # Global variables to maintain state
 model = None
-current_selections = []
-next_selection_id = 0
-active_selection_id = None
 
 class SelectionManager:
     def __init__(self):
@@ -62,14 +60,12 @@ def load_yolo_model():
     """Load YOLO model with error handling"""
     global model
     try:
-        # Using YOLOv8n for speed, can change to YOLOv8s, YOLOv8m, YOLOv8l, YOLOv8x for better accuracy
-        model = YOLO('yolov8n.pt')  # This will download the model on first run
+        model = YOLO('yolov8n.pt')
         return "✅ YOLO Model loaded successfully"
     except Exception as e:
         return f"❌ Error loading YOLO model: {str(e)}"
 
 def detect_objects(image, confidence_threshold=0.5, merge_with_existing=True):
-    """Detect objects using YOLO model"""
     global model, selection_manager
     
     if model is None:
@@ -79,31 +75,25 @@ def detect_objects(image, confidence_threshold=0.5, merge_with_existing=True):
         return None, "❌ No image provided", ""
     
     try:
-        # Convert PIL to numpy array for YOLO
         img_array = np.array(image)
         
-        # Run YOLO detection
         start_time = time.time()
         results = model(img_array, conf=confidence_threshold, verbose=False)
-        detection_time = (time.time() - start_time) * 1000  # Convert to ms
+        detection_time = (time.time() - start_time) * 1000
         
-        # Clear existing YOLO detections if not merging
         if not merge_with_existing:
             selection_manager.clear_type('yolo')
         
-        # Process detections
         detections_added = 0
         for result in results:
             boxes = result.boxes
             if boxes is not None:
                 for box in boxes:
-                    # Get bounding box coordinates
                     x1, y1, x2, y2 = box.xyxy[0].cpu().numpy()
                     confidence = box.conf[0].cpu().numpy()
                     class_id = int(box.cls[0].cpu().numpy())
                     class_name = model.names[class_id]
                     
-                    # Add selection
                     selection_manager.add_selection(
                         x=int(x1),
                         y=int(y1),
@@ -115,7 +105,6 @@ def detect_objects(image, confidence_threshold=0.5, merge_with_existing=True):
                     )
                     detections_added += 1
         
-        # Draw selections on image
         annotated_image = draw_selections(image)
         
         status_msg = f"✅ Detected {detections_added} objects in {detection_time:.1f}ms"
@@ -127,42 +116,34 @@ def detect_objects(image, confidence_threshold=0.5, merge_with_existing=True):
         return image, f"❌ Detection error: {str(e)}", get_analysis_text(image)
 
 def draw_selections(image):
-    """Draw all selections on the image"""
     if image is None:
         return None
     
-    # Create a copy of the image to draw on
     img_copy = image.copy()
     draw = ImageDraw.Draw(img_copy)
     
-    # Try to load a font, fallback to default if not available
     try:
         font = ImageFont.truetype("arial.ttf", 12)
     except:
         font = ImageFont.load_default()
     
-    # Draw each selection
     for selection in selection_manager.selections:
         x, y, w, h = selection['x'], selection['y'], selection['width'], selection['height']
         
-        # Choose color based on type
         if selection['type'] == 'yolo':
             outline_color = 'green'
-            fill_color = (0, 255, 0, 60)  # Semi-transparent green
+            fill_color = (0, 255, 0, 60)
         else:
             outline_color = 'blue'
-            fill_color = (0, 0, 255, 60)  # Semi-transparent blue
+            fill_color = (0, 0, 255, 60)
         
-        # Draw rectangle
         draw.rectangle([x, y, x + w, y + h], outline=outline_color, width=2)
         
-        # Draw semi-transparent fill
         overlay = Image.new('RGBA', img_copy.size, (0, 0, 0, 0))
         overlay_draw = ImageDraw.Draw(overlay)
         overlay_draw.rectangle([x, y, x + w, y + h], fill=fill_color)
         img_copy = Image.alpha_composite(img_copy.convert('RGBA'), overlay).convert('RGB')
         
-        # Draw label
         if selection['label']:
             label_y = y - 15 if y > 15 else y + h + 5
             draw.text((x, label_y), selection['label'], fill=outline_color, font=font)
@@ -170,7 +151,6 @@ def draw_selections(image):
     return img_copy
 
 def get_analysis_text(image):
-    """Generate analysis text for current selections"""
     if image is None:
         return "No image loaded"
     
@@ -187,11 +167,11 @@ def get_analysis_text(image):
     analysis = f"""
 ## 📊 Space Analysis
 
-**Total Image Area:** {total_image_area:,} px²
-**Total Selected Area:** {total_selected_area:,} px²
-**Selected Region:** {selected_percentage:.1f}% of total
-**Extra Space:** {extra_percentage:.1f}% of total
-**Number of Selections:** {len(selection_manager.selections)}
+**Total Image Area:** {total_image_area:,} px²  
+**Total Selected Area:** {total_selected_area:,} px²  
+**Selected Region:** {selected_percentage:.1f}% of total  
+**Extra Space:** {extra_percentage:.1f}% of total  
+**Number of Selections:** {len(selection_manager.selections)}  
 
 ### 📋 Selection Details:
 """
@@ -199,10 +179,10 @@ def get_analysis_text(image):
     for i, selection in enumerate(selection_manager.selections, 1):
         area = selection['width'] * selection['height']
         analysis += f"""
-**{i}.** {selection['label'] or f"Selection {selection['id']}"}
-- Type: {selection['type'].upper()}
-- Area: {area:,} px²
-- Dimensions: {selection['width']}×{selection['height']}
+**{i}.** {selection['label'] or f"Selection {selection['id']}"}  
+- Type: {selection['type'].upper()}  
+- Area: {area:,} px²  
+- Dimensions: {selection['width']}×{selection['height']}  
 """
         if selection['confidence']:
             analysis += f"- Confidence: {selection['confidence']:.2f}\n"
@@ -210,27 +190,22 @@ def get_analysis_text(image):
     return analysis
 
 def add_manual_selection(image, selection_data):
-    """Add manual selection from user input"""
     if image is None:
         return image, "❌ No image loaded", ""
     
     try:
-        # Parse selection data (format: "x,y,width,height")
         coords = [int(x.strip()) for x in selection_data.split(',')]
         if len(coords) != 4:
             raise ValueError("Invalid format")
         
         x, y, width, height = coords
         
-        # Validate coordinates
         img_width, img_height = image.size
         if x < 0 or y < 0 or x + width > img_width or y + height > img_height:
             return image, "❌ Selection coordinates out of bounds", get_analysis_text(image)
         
-        # Add selection
-        sel_id = selection_manager.add_selection(x, y, width, height, "manual", f"Manual {sel_id}")
+        selection_manager.add_selection(x, y, width, height, "manual", "Manual Selection")
         
-        # Redraw image
         annotated_image = draw_selections(image)
         analysis_text = get_analysis_text(image)
         
@@ -240,14 +215,12 @@ def add_manual_selection(image, selection_data):
         return image, f"❌ Error adding selection: {str(e)}", get_analysis_text(image)
 
 def clear_all_selections(image):
-    """Clear all selections"""
     selection_manager.clear_all()
     if image is not None:
         return image.copy(), "✅ All selections cleared", get_analysis_text(image)
     return None, "✅ All selections cleared", ""
 
 def clear_yolo_selections(image):
-    """Clear only YOLO detections"""
     selection_manager.clear_type('yolo')
     if image is not None:
         annotated_image = draw_selections(image)
@@ -255,7 +228,6 @@ def clear_yolo_selections(image):
     return None, "✅ YOLO detections cleared", ""
 
 def clear_manual_selections(image):
-    """Clear only manual selections"""
     selection_manager.clear_type('manual')
     if image is not None:
         annotated_image = draw_selections(image)
@@ -263,205 +235,62 @@ def clear_manual_selections(image):
     return None, "✅ Manual selections cleared", ""
 
 def process_image_upload(image):
-    """Process uploaded image"""
     if image is None:
         return None, "❌ No image uploaded", ""
     
-    # Clear previous selections
     selection_manager.clear_all()
-    
-    # Return original image and analysis
     analysis_text = get_analysis_text(image)
     return image, "✅ Image loaded successfully", analysis_text
 
-# Custom CSS for better styling
-css = """
-.gradio-container {
-    max-width: 1400px !important;
-}
-.analysis-text {
-    font-family: 'Courier New', monospace;
-    background-color: #f8f9fa;
-    padding: 15px;
-    border-radius: 8px;
-    border-left: 4px solid #007bff;
-}
-"""
-
-# Create Gradio interface
+# Gradio Interface creation (exported as app)
 def create_interface():
+    css = """
+    .gradio-container {
+        max-width: 1400px !important;
+    }
+    .analysis-text {
+        font-family: 'Courier New', monospace;
+        background-color: #f8f9fa;
+        padding: 15px;
+        border-radius: 8px;
+        border-left: 4px solid #007bff;
+    }
+    """
+    
     with gr.Blocks(css=css, title="Image Space Analyzer with YOLO") as demo:
-        gr.Markdown("""
-        # 🔍 Image Space Analyzer with YOLO
-        
-        **Fast object detection and space analysis tool**
-        
-        Upload an image to start analyzing space usage with automatic YOLO detection or manual selections.
-        """)
+        gr.Markdown("# 🔍 Image Space Analyzer with YOLO")
         
         with gr.Row():
             with gr.Column(scale=2):
-                # Image input/output
-                image_input = gr.Image(
-                    type="pil",
-                    label="📸 Upload Image",
-                    height=500
-                )
-                
-                # Status display
-                status_output = gr.Textbox(
-                    label="📋 Status",
-                    interactive=False,
-                    max_lines=2
-                )
-                
+                image_input = gr.Image(type="pil", label="📸 Upload Image", height=500)
+                status_output = gr.Textbox(label="📋 Status", interactive=False, max_lines=2)
             with gr.Column(scale=1):
-                # Model status
-                model_status = gr.Textbox(
-                    label="🤖 Model Status",
-                    value="Loading YOLO model...",
-                    interactive=False
-                )
-                
-                # YOLO Detection Controls
+                model_status = gr.Textbox(label="🤖 Model Status", value="Loading YOLO model...", interactive=False)
                 gr.Markdown("### 🎯 YOLO Object Detection")
-                
-                with gr.Row():
-                    confidence_slider = gr.Slider(
-                        minimum=0.1,
-                        maximum=1.0,
-                        value=0.5,
-                        step=0.1,
-                        label="Confidence Threshold"
-                    )
-                
-                merge_checkbox = gr.Checkbox(
-                    label="Merge with existing selections",
-                    value=True
-                )
-                
+                confidence_slider = gr.Slider(minimum=0.1, maximum=1.0, value=0.5, step=0.1, label="Confidence Threshold")
+                merge_checkbox = gr.Checkbox(label="Merge with existing selections", value=True)
                 detect_btn = gr.Button("🔍 Detect Objects", variant="primary")
-                
-                # Manual Selection Controls
                 gr.Markdown("### ✏️ Manual Selection")
-                
-                manual_input = gr.Textbox(
-                    label="Selection (x,y,width,height)",
-                    placeholder="100,100,200,150",
-                    info="Enter coordinates separated by commas"
-                )
-                
+                manual_input = gr.Textbox(label="Selection (x,y,width,height)", placeholder="100,100,200,150")
                 add_manual_btn = gr.Button("➕ Add Manual Selection")
-                
-                # Selection Management
                 gr.Markdown("### 🗂️ Selection Management")
-                
-                with gr.Row():
-                    clear_all_btn = gr.Button("🗑️ Clear All", variant="secondary")
-                    clear_yolo_btn = gr.Button("🗑️ Clear YOLO")
-                    clear_manual_btn = gr.Button("🗑️ Clear Manual")
+                clear_all_btn = gr.Button("🗑️ Clear All", variant="secondary")
+                clear_yolo_btn = gr.Button("🗑️ Clear YOLO")
+                clear_manual_btn = gr.Button("🗑️ Clear Manual")
         
-        # Analysis output
         with gr.Row():
-            analysis_output = gr.Markdown(
-                label="📊 Analysis Results",
-                elem_classes=["analysis-text"]
-            )
-        
-        # Event handlers
-        image_input.upload(
-            fn=process_image_upload,
-            inputs=[image_input],
-            outputs=[image_input, status_output, analysis_output]
-        )
-        
-        detect_btn.click(
-            fn=detect_objects,
-            inputs=[image_input, confidence_slider, merge_checkbox],
-            outputs=[image_input, status_output, analysis_output]
-        )
-        
-        add_manual_btn.click(
-            fn=add_manual_selection,
-            inputs=[image_input, manual_input],
-            outputs=[image_input, status_output, analysis_output]
-        )
+            analysis_output = gr.Markdown(label="📊 Analysis Results", elem_classes=["analysis-text"])
         
-        clear_all_btn.click(
-            fn=clear_all_selections,
-            inputs=[image_input],
-            outputs=[image_input, status_output, analysis_output]
-        )
+        image_input.upload(process_image_upload, [image_input], [image_input, status_output, analysis_output])
+        detect_btn.click(detect_objects, [image_input, confidence_slider, merge_checkbox], [image_input, status_output, analysis_output])
+        add_manual_btn.click(add_manual_selection, [image_input, manual_input], [image_input, status_output, analysis_output])
+        clear_all_btn.click(clear_all_selections, [image_input], [image_input, status_output, analysis_output])
+        clear_yolo_btn.click(clear_yolo_selections, [image_input], [image_input, status_output, analysis_output])
+        clear_manual_btn.click(clear_manual_selections, [image_input], [image_input, status_output, analysis_output])
         
-        clear_yolo_btn.click(
-            fn=clear_yolo_selections,
-            inputs=[image_input],
-            outputs=[image_input, status_output, analysis_output]
-        )
-        
-        clear_manual_btn.click(
-            fn=clear_manual_selections,
-            inputs=[image_input],
-            outputs=[image_input, status_output, analysis_output]
-        )
-        
-        # Load model on startup
-        demo.load(
-            fn=load_yolo_model,
-            outputs=[model_status]
-        )
-        
-        gr.Markdown("""
-        ## 📖 How to Use:
-        
-        1. **Upload an image** using the image upload area
-        2. **Adjust confidence threshold** for YOLO detection sensitivity
-        3. **Click "Detect Objects"** to automatically detect objects using YOLO
-        4. **Add manual selections** by entering coordinates (x,y,width,height)
-        5. **View analysis results** showing space usage statistics
-        6. **Manage selections** using the clear buttons
-        
-        ### 🔧 Features:
-        - ⚡ **Ultra-fast YOLO detection** (millisecond response times)
-        - 🎯 **High accuracy object detection** with confidence scores
-        - ✏️ **Manual selection support** for custom regions
-        - 📊 **Detailed space analysis** with percentages and areas
-        - 🗂️ **Selection management** with type-based filtering
-        - 🎨 **Visual feedback** with color-coded selections
-        
-        ### 🚀 Performance:
-        - Uses YOLOv8n for optimal speed/accuracy balance
-        - GPU acceleration when available
-        - Real-time analysis updates
-        """)
+        demo.load(load_yolo_model, outputs=[model_status])
     
     return demo
 
-# Additional requirements.txt content that should be installed:
-requirements = """
-gradio>=4.0.0
-ultralytics>=8.0.0
-torch>=2.0.0
-torchvision>=0.15.0
-opencv-python>=4.8.0
-numpy>=1.24.0
-Pillow>=10.0.0
-"""
-
-if __name__ == "__main__":
-    # Print requirements
-    print("📋 Required packages:")
-    print(requirements)
-    print("\n" + "="*50)
-    print("🚀 Starting Image Space Analyzer with YOLO...")
-    print("="*50)
-    
-    # Create and launch the interface
-    demo = create_interface()
-    demo.launch(
-        server_name="0.0.0.0",  # Allow external access
-        server_port=7860,       # Default Gradio port
-        share=False,            # Set to True to create public link
-        debug=True,
-        show_error=True
-    )
+# Export app (required for Hugging Face Spaces!)
+app = create_interface()