app.py

import gradio as gr
from ultralyticsplus import YOLO, render_result
import cv2

# Load the YOLO model
model = YOLO('foduucom/plant-leaf-detection-and-classification')

# Set model parameters
model.overrides['conf'] = 0.25  # NMS confidence threshold
model.overrides['iou'] = 0.45  # NMS IoU threshold
model.overrides['agnostic_nms'] = False  # NMS class-agnostic
model.overrides['max_det'] = 1000  # maximum detections per image

def detect_leaves(image):
    # Convert from Gradio's numpy array to image file
    image = cv2.cvtColor(image, cv2.COLOR_RGB2BGR)
    cv2.imwrite('temp_image.jpg', image)
    
    # Perform prediction
    results = model.predict('temp_image.jpg')
    
    # Count number of leaves detected
    num_leaves = len(results[0].boxes)
    
    # Render results on image
    render = render_result(model=model, image='temp_image.jpg', result=results[0])
    
    return render, num_leaves

# Create Gradio interface
with gr.Blocks(title="Leaf Detection & Classification") as demo:
    gr.Markdown("# 🍃 Plant Leaf Detection & Classification")
    gr.Markdown("Upload a plant image to detect and count leaves")
    
    with gr.Row():
        with gr.Column():
            input_image = gr.Image(label="Upload Plant Image")
            submit_btn = gr.Button("Detect Leaves")
        
        with gr.Column():
            output_image = gr.Image(label="Detection Results")
            leaf_count = gr.Number(label="Number of Leaves Detected")
    
    submit_btn.click(
        fn=detect_leaves,
        inputs=[input_image],
        outputs=[output_image, leaf_count]
    )

if __name__ == "__main__":
    demo.launch(server_port=7860, share=False)