app.py

import cv2
import torch
import numpy as np
from PIL import Image
from torchvision import models, transforms
from ultralytics import YOLO
import gradio as gr
import torch.nn as nn

# Initialize device
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

# Load models
yolo_model = YOLO('best.pt')  # Make sure this file is uploaded to your Space
resnet = models.resnet50(pretrained=False)

# Modify ResNet for 3 classes
resnet.fc = nn.Linear(resnet.fc.in_features, 3)
resnet.load_state_dict(torch.load('rice_resnet_model.pth', map_location=device))
resnet = resnet.to(device)
resnet.eval()

# Class labels
class_labels = ["c9", "kant", "superf"]

# Image transformations
transform = transforms.Compose([
    transforms.Resize((224, 224)),
    transforms.ToTensor(),
    transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
])

def classify_crop(crop_img):
    """Classify a single rice grain"""
    image = transform(crop_img).unsqueeze(0).to(device)
    with torch.no_grad():
        output = resnet(image)
        _, predicted = torch.max(output, 1)
    return class_labels[predicted.item()]

def detect_and_classify(image):
    """Process full image with YOLO + ResNet"""
    image = np.array(image)
    results = yolo_model(image)[0]
    boxes = results.boxes.xyxy.cpu().numpy()

    for box in boxes:
        x1, y1, x2, y2 = map(int, box[:4])
        crop = image[y1:y2, x1:x2]
        crop_pil = Image.fromarray(cv2.cvtColor(crop, cv2.COLOR_BGR2RGB))
        predicted_label = classify_crop(crop_pil)

        # Draw bounding box and label
        cv2.rectangle(image, (x1, y1), (x2, y2), (0, 255, 0), 2)
        cv2.putText(image, predicted_label, (x1, y1-10), 
                   cv2.FONT_HERSHEY_SIMPLEX, 0.9, (36, 255, 12), 2)

    return Image.fromarray(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))

# Gradio Interface
with gr.Blocks(title="چاول کا شناختی نظام") as demo:
    gr.Markdown("""
    # چاول کا شناختی نظام
    ایک تصویر اپ لوڈ کریں جس میں چاول کے دانے ہوں۔ نظام ہر دانے کو پہچان کر اس کی قسم بتائے گا۔
    """)
    
    with gr.Row():
        input_image = gr.Image(type="pil", label="تصویر داخل کریں")
        output_image = gr.Image(type="pil", label="نتیجہ")
    
    submit_btn = gr.Button("تشخیص کریں")
    submit_btn.click(
        fn=detect_and_classify,
        inputs=input_image,
        outputs=output_image
    )
    
    gr.Examples(
        examples=[["example1.jpg"], ["example2.jpg"]],  # Add your example images
        inputs=input_image,
        outputs=output_image,
        fn=detect_and_classify,
        cache_examples=True
    )

demo.launch()