Update app.py

app.py

@@ -26,6 +26,10 @@ class GradioLettuceAnalysisPipeline:
         self.growth_model_path = growth_model_path
         self.health_classification_model_path = health_classification_model_path
         
+        # Fixed confidence thresholds (no longer adjustable via sliders)
+        self.detection_confidence = 0.5
+        self.growth_confidence = 0.25
+        
         # Load all models
         self.load_models()
         
@@ -48,10 +52,10 @@ class GradioLettuceAnalysisPipeline:
             # 3. Load health classification model
             self.load_health_classification_model()
             
-            return "All models loaded successfully!"
+            return "✅ All models loaded successfully!"
             
         except Exception as e:
-            return f"Error loading models: {e}"
+            return f"❌ Error loading models: {e}"
     
     def load_health_classification_model(self):
         """Load the health classification model (ViT)"""
@@ -68,9 +72,9 @@ class GradioLettuceAnalysisPipeline:
         self.health_classification_model.load_state_dict(checkpoint['model_state_dict'])
         self.health_classification_model.eval()
     
-    def detect_lettuce(self, image_path, confidence_threshold=0.5):
+    def detect_lettuce(self, image_path):
         """Stage 1: Detect lettuce in the image"""
-        results = self.detection_model(image_path, conf=confidence_threshold)
+        results = self.detection_model(image_path, conf=self.detection_confidence)
         detections = []
         
         for result in results:
@@ -90,7 +94,7 @@ class GradioLettuceAnalysisPipeline:
         
         return detections
     
-    def classify_growth_stage(self, image_path, bbox, confidence_threshold=0.25):
+    def classify_growth_stage(self, image_path, bbox):
         """Stage 2: Classify growth stage"""
         try:
             image = Image.open(image_path)
@@ -111,7 +115,7 @@ class GradioLettuceAnalysisPipeline:
             # Run growth stage classification
             results = self.growth_model.predict(
                 source=temp_crop_path,
-                conf=confidence_threshold,
+                conf=self.growth_confidence,
                 save=False,
                 imgsz=640,
                 verbose=False
@@ -165,12 +169,12 @@ class GradioLettuceAnalysisPipeline:
         except Exception as e:
             return "Unknown", 0.0
     
-    def process_image_gradio(self, image, detection_conf, growth_conf, show_boxes, show_labels):
+    def process_image_gradio(self, image, show_boxes, show_labels):
         """
         Process image for Gradio interface
         """
         if image is None:
-            return None, "❌ Please upload an image first!", None, None
+            return None, "Please upload an image first!", None, None
         
         try:
             # Save uploaded image temporarily
@@ -182,13 +186,13 @@ class GradioLettuceAnalysisPipeline:
                 image = image.convert('RGB')
             
             # Stage 1: Detect lettuce
-            detections = self.detect_lettuce(temp_image_path, detection_conf)
+            detections = self.detect_lettuce(temp_image_path)
             
             if not detections:
                 # Clean up
                 if os.path.exists(temp_image_path):
                     os.remove(temp_image_path)
-                return image, "❌ No lettuce detected in the image!", None, None
+                return image, "No lettuce detected in the image!", None, None
             
             # Process each detection
             complete_results = []
@@ -210,7 +214,7 @@ class GradioLettuceAnalysisPipeline:
                 det_conf = detection['confidence']
                 
                 # Stage 2: Growth stage
-                growth_stage, growth_conf = self.classify_growth_stage(temp_image_path, bbox, growth_conf)
+                growth_stage, growth_conf = self.classify_growth_stage(temp_image_path, bbox)
                 
                 # Stage 3: Health status
                 health_status, health_conf = self.classify_health(image, bbox)
@@ -292,7 +296,7 @@ class GradioLettuceAnalysisPipeline:
         if not results:
             return "No results to display"
         
-        summary = f"🔍 **LETTUCE ANALYSIS RESULTS**\n\n"
+        summary = f"**LETTUCE ANALYSIS RESULTS**\n\n"
         summary += f"**Summary:**\n"
         summary += f"- Total lettuce detected: **{len(results)}**\n"
         
@@ -307,7 +311,7 @@ class GradioLettuceAnalysisPipeline:
         summary += f"- Health statuses: {dict(health_counts)}\n\n"
         
         # Detailed results
-        summary += f"**Detailed Results:**\n\n"
+        summary += f"📋 **Detailed Results:**\n\n"
         
         for result in results:
             summary += f"**Lettuce {result['lettuce_id']}:**\n"
@@ -348,12 +352,12 @@ except Exception as e:
     model_status = f"Error loading models: {e}"
     pipeline = None
 
-def process_image_wrapper(image, detection_conf, growth_conf, show_boxes, show_labels):
+def process_image_wrapper(image, show_boxes, show_labels):
     """Wrapper function for Gradio interface"""
     if pipeline is None:
         return None, "Models not loaded properly!", None
     
-    return pipeline.process_image_gradio(image, detection_conf, growth_conf, show_boxes, show_labels)
+    return pipeline.process_image_gradio(image, show_boxes, show_labels)
 
 def download_results(results):
     """Create downloadable results"""
@@ -388,84 +392,127 @@ def download_results(results):
     
     return filename
 
+# Custom CSS for styling and logo
+custom_css = """
+.logo-container {
+    text-align: center;
+    margin-bottom: 20px;
+}
+
+.logo-container img {
+    max-height: 100px;
+    width: auto;
+}
+
+.company-header {
+    text-align: center;
+    margin-bottom: 30px;
+    padding: 20px;
+    background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
+    border-radius: 10px;
+    color: white;
+}
+
+.analyze-button {
+    background: linear-gradient(45deg, #4CAF50, #45a049) !important;
+    color: white !important;
+    border: none !important;
+    padding: 15px 30px !important;
+    font-size: 16px !important;
+    font-weight: bold !important;
+    border-radius: 8px !important;
+    cursor: pointer !important;
+    transition: all 0.3s ease !important;
+}
+
+.analyze-button:hover {
+    transform: translateY(-2px) !important;
+    box-shadow: 0 4px 8px rgba(0,0,0,0.2) !important;
+}
+
+.settings-container {
+    background: #f8f9fa;
+    padding: 20px;
+    border-radius: 10px;
+    margin-bottom: 20px;
+}
+
+.footer-info {
+    background: #f1f3f4;
+    padding: 20px;
+    border-radius: 10px;
+    margin-top: 20px;
+}
+"""
+
 # Create Gradio interface
-with gr.Blocks(title="Lettuce Analysis Pipeline", theme=gr.themes.Soft()) as demo:
-    
-    # Header
-    gr.Markdown("""
-    # Complete Lettuce Analysis Pipeline
+with gr.Blocks(title="Lettuce Analysis Pipeline", theme=gr.themes.Soft(), css=custom_css) as demo:
     
-    This application performs comprehensive lettuce analysis in three stages:
-    1. **Detection**: Locates lettuce in your image
-    2. **Growth Stage**: Classifies the growth stage of each lettuce
-    3. **Health Status**: Determines the health condition of each lettuce
+    # Company Header with Logo
+    with gr.Row():
+        gr.HTML("""
+                <div class="company-header"><div class="logo-container">
+                <img src="./GB_logo.jpg" alt="Garden Of Babylon" />
+                </div> 
+                <h1>Advanced Lettuce Analysis Platform</h1>
+                    <p>Powered by AI • Precision Agriculture Solutions</p>
+                </div> 
+                """)
     
-    Upload an image and adjust the settings to get started!
-    """)
+    # Main content
+    # gr.Markdown("""## Professional Lettuce Analysis Pipeline Our advanced AI system performs comprehensive lettuce analysis in three automated stages: - **Detection**: Automatically locates lettuce in your images - **Growth Stage Classification**: Determines the growth stage of each lettuce plant- **Health Assessment**: Evaluates the health condition of each plant Simply upload an image and let our AI do the rest!""")
     
     # Model status
-    gr.Markdown(f"**Model Status:** {model_status}")
+    gr.Markdown(f"**System Status:** {model_status}")
     
     with gr.Row():
         # Left column - Input
         with gr.Column(scale=1):
-            gr.Markdown("## Input")
+            gr.Markdown("## Upload Image")
             
             # Image input
             input_image = gr.Image(
                 type="pil",
                 label="Upload Lettuce Image",
                 sources=["upload"],
-                interactive=True
+                interactive=True,
+                height=300
             )
             
-            # # Settings
-            # gr.Markdown("### Settings")
-            
-            # detection_conf = gr.Slider(
-            #     minimum=0.1,
-            #     maximum=1.0,
-            #     value=0.5,
-            #     step=0.05,
-            #     label="Detection Confidence Threshold",
-            #     info="Higher values = more strict detection"
-            # )
-            
-            # growth_conf = gr.Slider(
-            #     minimum=0.1,
-            #     maximum=1.0,
-            #     value=0.25,
-            #     step=0.05,
-            #     label="Growth Stage Confidence Threshold",
-            #     info="Higher values = more strict growth classification"
-            # )
-            
-            with gr.Row():
-                show_boxes = gr.Checkbox(
-                    label="Show Bounding Boxes",
-                    value=True
-                )
-                show_labels = gr.Checkbox(
-                    label="Show Labels",
-                    value=True
-                )
+            # Simplified settings
+            with gr.Group():
+                gr.Markdown("### Display Options")
+                with gr.Row():
+                    show_boxes = gr.Checkbox(
+                        label="Show Bounding Boxes",
+                        value=True
+                    )
+                    show_labels = gr.Checkbox(
+                        label="Show Labels",
+                        value=True
+                    )
             
             # Process button
             process_btn = gr.Button(
-                "Analyze Lettuce",
+                "🚀 Analyze Lettuce",
                 variant="primary",
-                size="lg"
+                size="lg",
+                elem_classes="analyze-button"
             )
+            
+            # Info box
+            #gr.Markdown("""<div class="settings-container"><h4>ℹ️ Analysis Settings</h4><ul><li><strong>Detection Confidence:</strong> Optimized at 50%</li><li><strong>Growth Classification:</strong> Optimized at 25%</li><li><strong>Processing Time:</strong> ~5-15 seconds per image</li></ul></div>""")
         
         # Right column - Output
         with gr.Column(scale=2):
-            gr.Markdown("## Results")
+            gr.Markdown("## Analysis Results")
             
             # Output image
             output_image = gr.Image(
                 label="Analysis Results",
                 type="pil",
-                interactive=False
+                interactive=False,
+                height=400
             )
             
             # Results summary
@@ -475,17 +522,19 @@ with gr.Blocks(title="Lettuce Analysis Pipeline", theme=gr.themes.Soft()) as dem
             )
     
     # Results table
-    gr.Markdown("## Detailed Results Table")
+    gr.Markdown("##Detailed Results")
     results_table = gr.Dataframe(
-        label="Detailed Results",
+        label="Comprehensive Analysis Data",
         interactive=False,
         wrap=True
     )
     
     # Download section
     with gr.Row():
-        download_btn = gr.Button("Download Results (JSON)", variant="secondary")
-        download_file = gr.File(label="Download Results", visible=False)
+        with gr.Column(scale=1):
+            download_btn = gr.Button("Download Results (JSON)", variant="secondary")
+        with gr.Column(scale=2):
+            download_file = gr.File(label="Download Analysis Report", visible=False)
     
     # Hidden state to store results
     results_state = gr.State()
@@ -493,7 +542,7 @@ with gr.Blocks(title="Lettuce Analysis Pipeline", theme=gr.themes.Soft()) as dem
     # Event handlers
     process_btn.click(
         fn=process_image_wrapper,
-        inputs=[input_image, detection_conf, growth_conf, show_boxes, show_labels],
+        inputs=[input_image, show_boxes, show_labels],
         outputs=[output_image, results_summary, results_table, results_state]
     )
     
@@ -507,32 +556,36 @@ with gr.Blocks(title="Lettuce Analysis Pipeline", theme=gr.themes.Soft()) as dem
     )
     
     # Footer
-    
-    # ### Notes:
-    # - **Detection Confidence**: Lower values detect more lettuce but may include false positives
-    # - **Growth Confidence**: Lower values are more permissive for growth stage classification
-    # - **Bounding Boxes**: Toggle to show/hide detection boxes around lettuce
-    # - **Labels**: Toggle to show/hide detailed information labels
-    gr.Markdown("""
-    ---
-    ### Supported Features:
-    - Multiple lettuce detection in single image
-    - Real-time confidence adjustment
-    - Downloadable results in JSON format
-    - Detailed analysis summary and statistics
-    
-    **Developed with for precision agriculture**
+    gr.HTML("""
+    <div class="footer-info">
+        <h3>🔧 System Features</h3>
+        <div style="display: grid; grid-template-columns: repeat(auto-fit, minmax(250px, 1fr)); gap: 20px;">
+            <div>
+                <h4>Precision Detection</h4>
+                <p>Advanced YOLO-based detection with optimized confidence thresholds</p>
+            </div>
+            <div>
+                <h4> Growth Analysis</h4>
+                <p>Multi-stage classification for accurate growth assessment</p>
+            </div>
+            <div>
+                <h4>Health Monitoring</h4>
+                <p>Vision Transformer (ViT) powered health status evaluation</p>
+            </div>
+            <div>
+                <h4>Comprehensive Reports</h4>
+                <p>Detailed analysis with downloadable JSON reports</p>
+            </div>
+        </div>
+        <hr style="margin: 20px 0;">
+        <p style="text-align: center; color: #666;">
+            <strong>Developed for Precision Agriculture</strong> | 
+            Optimized confidence thresholds for maximum accuracy | 
+            Support for multiple lettuce detection
+        </p>
+    </div>
     """)
 
 # Launch the app
-demo.launch()
-
-"""
 if __name__ == "__main__":
-    demo.launch(
-        share=True,
-        server_name="0.0.0.0",
-        server_port=7860,
-        debug=True
-    )
-"""    
+    demo.launch()