import gradio as gr
from detection import ObjectDetection

examples = [
    ['test-images/plant1.jpeg', 0.23],
    ['test-images/plant2.jpeg', 0.45],
    ['test-images/plant3.webp', 0.43],
]

def get_predictions(img, threshold, box_color, text_color):
    v8_results = yolov8_detector.v8_score_frame(img)
    v8_frame = yolov8_detector.plot_bboxes(v8_results, img, float(threshold), box_color, text_color)
    return v8_frame

# Load the YOLOv8s model for plant leaf detection and classification
yolov8_detector = ObjectDetection('Yolov8')

with gr.Blocks(title="Plant Leaf Detection and Classification", theme=gr.themes.Soft()) as interface:
    gr.Markdown("# Plant Leaf Detection and Classification🍃")
    gr.Markdown("This app uses YOLOv8s, a powerful object detection model, to detect and classify plant leaves. "
                "The model can detect various plant leaves and classify them into up to 45 different plant categories, "
                "whether the leafs are  diseased or healthy.")
    gr.Markdown("If you like this app, don't forget to give it a thumbs up on Hugging Face!😊❤️@ www.Foduu.com")

    with gr.Row():
        with gr.Column():
            image = gr.Image(label="Input Image")
        with gr.Column():
            with gr.Row():
                with gr.Column():
                    box_color = gr.ColorPicker(label="Box Color", value="#FF8C00")
                with gr.Column():
                    text_color = gr.ColorPicker(label="Prediction Color", value="#000000")

            confidence = gr.Slider(maximum=1, step=0.01, value=0.6, label="Confidence Threshold", interactive=True)
            btn = gr.Button("Detect")
    
    with gr.Row():        
        with gr.Box():
            v8_prediction = gr.Image(label="Leaf Detection and Classification",height=100,width=100)  # Display the output image in its original size

        btn.click(
            get_predictions,
            [image, confidence, box_color, text_color],
            [v8_prediction]
        )
    
    with gr.Row():
        gr.Examples(examples=examples, inputs=[image, confidence])

interface.launch()