app.py

import cv2
import gradio as gr
import uuid
import os
import torch
from ultralytics import YOLO

# Download model if not present
model_path ="best.pt"

def detect(image):
    results = model(image)
    annotated_image = results[0].plot()  # draw boxes
    return annotated_image


def detect_faults(video_path):
        # Generate unique run directory
        unique_id = str(uuid.uuid4())[:8]
        output_dir = os.path.join("runs", "detect", unique_id)
        os.makedirs(output_dir, exist_ok=True)
        output_path = os.path.join(output_dir, "output.mp4")

        # Run YOLO prediction
        model.predict(
            source=video_path,
            save=True,
            save_txt=False,
            conf=0.5,
            project="runs/detect",
            name=unique_id
        )

        # Open the input video
        cap = cv2.VideoCapture(video_path)
        if not cap.isOpened():
            return "Error: Could not open video."

        fps = cap.get(cv2.CAP_PROP_FPS)
        width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
        height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))

        fourcc = cv2.VideoWriter_fourcc(*'mp4v')
        out = cv2.VideoWriter(output_path, fourcc, fps, (width, height))

        while True:
            ret, frame = cap.read()
            if not ret:
                break

            results = model(frame)
            annotated_frame = results[0].plot()
            out.write(annotated_frame)

        cap.release()
        out.release()

        
def detect_faults_in_image(image):
    try:
        results = model(image)
        annotated = results[0].plot()
        return annotated
    except Exception as e:
        print(f"❌ Error processing image: {e}")
        return "Image processing error."

# Gradio Interface
video_ui = gr.Interface(
    fn=detect_faults,
    inputs=gr.Video(label="Upload Drone Video (MP4 only)"),
    outputs=gr.Video(label="Detected Faults Video"),
    title="Solar Panel Fault Detection (Video)",
    description="Upload a drone video of solar panels. The model detects faults and returns an annotated video."
)

# Gradio Interface
demo = gr.Interface(
    fn=detect,
    inputs=gr.Image(type="filepath", label="Upload an Image"),
    outputs=gr.Image(type="numpy", label="Detected Output"),
    title="Solar Panel Fault Detector - YOLOv8",
    description="Upload a drone image of solar panels. YOLOv8 will detect any faults."
)

demo.launch()


app = gr.TabbedInterface([image_ui, video_ui], ["Image Detection"], "Video Detection")

if __name__ == "__main__":
    app.launch()