Add application file

app.py

@@ -0,0 +1,146 @@
+import gradio as gr
+from ultralytics import YOLO
+import cv2
+import numpy as np
+from PIL import Image, ImageDraw, ImageFont
+import requests
+from io import BytesIO
+
+# Load the model
+def load_model():
+    # Replace with your model path on HuggingFace
+    model = YOLO('https://huggingface.co/IsmatS/crop_desease_detection/resolve/main/best.pt')
+    return model
+
+model = load_model()
+
+def detect_tree_disease(image, conf_threshold=0.25, iou_threshold=0.45):
+    """Detect unhealthy trees in the uploaded image"""
+    
+    # Convert PIL image to numpy array
+    image_np = np.array(image)
+    
+    # Run inference
+    results = model(image_np, conf=conf_threshold, iou=iou_threshold)
+    
+    # Extract detections
+    detections = []
+    for result in results:
+        boxes = result.boxes
+        if boxes is not None:
+            for box in boxes:
+                detection = {
+                    'confidence': float(box.conf[0]),
+                    'bbox': box.xyxy[0].tolist(),
+                    'class': 'unhealthy'
+                }
+                detections.append(detection)
+    
+    # Draw annotations on the image
+    annotated_img = image.copy()
+    draw = ImageDraw.Draw(annotated_img)
+    
+    # Try to use a default font, fall back to PIL default if not available
+    try:
+        font = ImageFont.truetype("arial.ttf", 20)
+    except:
+        font = ImageFont.load_default()
+    
+    for det in detections:
+        bbox = det['bbox']
+        conf = det['confidence']
+        
+        # Draw bounding box
+        draw.rectangle(bbox, outline="red", width=3)
+        
+        # Draw label with confidence
+        label = f"Unhealthy: {conf:.2f}"
+        text_bbox = draw.textbbox((bbox[0], bbox[1] - 25), label, font=font)
+        draw.rectangle(text_bbox, fill="red")
+        draw.text((bbox[0], bbox[1] - 25), label, fill="white", font=font)
+    
+    # Create detection summary
+    summary = f"Detected {len(detections)} unhealthy tree(s)\n\n"
+    for i, det in enumerate(detections, 1):
+        summary += f"Tree {i}: Confidence {det['confidence']:.2f}\n"
+    
+    return annotated_img, summary
+
+# Create example images
+example_images = [
+    ["https://raw.githubusercontent.com/ultralytics/yolov5/master/data/images/zidane.jpg", 0.25, 0.45],
+    # Add your own example images here
+]
+
+# Create Gradio interface
+with gr.Blocks(title="Tree Disease Detection") as demo:
+    gr.Markdown("""
+    # 🌳 Tree Disease Detection with YOLOv8
+    
+    This model detects unhealthy/diseased trees in aerial UAV imagery. 
+    Upload an image or use one of the examples below to detect diseased trees.
+    
+    **Model**: YOLOv8s trained on PDT (Pests and Diseases Tree) dataset
+    """)
+    
+    with gr.Row():
+        with gr.Column():
+            input_image = gr.Image(type="pil", label="Upload Image")
+            conf_threshold = gr.Slider(
+                minimum=0.0, 
+                maximum=1.0, 
+                value=0.25, 
+                step=0.05,
+                label="Confidence Threshold"
+            )
+            iou_threshold = gr.Slider(
+                minimum=0.0, 
+                maximum=1.0, 
+                value=0.45, 
+                step=0.05,
+                label="IoU Threshold"
+            )
+            detect_button = gr.Button("Detect Tree Disease")
+        
+        with gr.Column():
+            output_image = gr.Image(type="pil", label="Detection Results")
+            detection_summary = gr.Textbox(label="Detection Summary", lines=10)
+    
+    # Set up event handler
+    detect_button.click(
+        fn=detect_tree_disease,
+        inputs=[input_image, conf_threshold, iou_threshold],
+        outputs=[output_image, detection_summary]
+    )
+    
+    # Add examples
+    gr.Examples(
+        examples=example_images,
+        inputs=[input_image, conf_threshold, iou_threshold],
+        outputs=[output_image, detection_summary],
+        fn=detect_tree_disease,
+        cache_examples=True,
+    )
+    
+    gr.Markdown("""
+    ## About this Model
+    
+    - **Architecture**: YOLOv8s
+    - **Dataset**: [PDT Dataset](https://huggingface.co/datasets/qwer0213/PDT_dataset)
+    - **mAP50**: 0.933
+    - **mAP50-95**: 0.659
+    - **Classes**: 1 (unhealthy trees)
+    
+    ## Usage Tips
+    
+    - This model works best with aerial/UAV imagery
+    - Optimal input resolution: 640x640 pixels
+    - Adjust confidence threshold to filter detections
+    - Lower IoU threshold for overlapping trees
+    
+    [Model Card](https://huggingface.co/IsmatS/crop_desease_detection) | 
+    [Dataset](https://huggingface.co/datasets/qwer0213/PDT_dataset)
+    """)
+
+# Launch the app
+demo.launch()

requirements.txt

@@ -0,0 +1,5 @@
+gradio
+ultralytics
+opencv-python-headless
+pillow
+numpy