app.py

from flask import Flask, request, send_file, jsonify
from flask_cors import CORS
import numpy as np
from PIL import Image
import io
import torch
from basicsr.archs.rrdbnet_arch import RRDBNet
from realesrgan import RealESRGANer
import time
import os

app = Flask(__name__)
CORS(app)

# Global variables
upsampler = None

# Initialize Real-ESRGAN upsampler at startup
def initialize_enhancer():
    global upsampler
    
    try:
        print("Starting model initialization...")
        # Configuration for RealESRGAN x4v3
        model = RRDBNet(
            num_in_ch=3,
            num_out_ch=3,
            num_feat=64,
            num_block=6,  # Critical parameter for x4v3 model
            num_grow_ch=32,
            scale=4
        )
        
        # Force CPU usage for Hugging Face compatibility
        device = torch.device('cpu')
        
        # Check if model weights file exists
        weights_path = 'weights/realesr-general-x4v3.pth'
        if not os.path.exists(weights_path):
            print(f"Model weights not found at {weights_path}")
            # Create the directory if it doesn't exist
            os.makedirs('weights', exist_ok=True)
            # You'd normally download weights here, but for this example
            # we'll just use a placeholder
            print("ERROR: Model weights file not found!")
            return False
        
        # Initialize the upsampler
        upsampler = RealESRGANer(
            scale=4,
            model_path=weights_path,
            model=model,
            tile=0,  # Set to 0 for small images, increase for large images
            tile_pad=10,
            pre_pad=0,
            half=False,  # CPU doesn't support half precision
            device=device
        )
        
        # Run a small test to ensure everything is loaded
        test_img = np.zeros((64, 64, 3), dtype=np.uint8)
        upsampler.enhance(test_img, outscale=4, alpha_upsampler='realesrgan')
        
        print("Model initialization completed successfully")
        return True
        
    except Exception as e:
        error_msg = f"Model initialization failed: {str(e)}"
        print(error_msg)
        return False

# Global init flag to track if we've attempted initialization
init_attempted = False

# Initialize model immediately at startup - this blocks until model is ready
if not initialize_enhancer():
    print("ERROR: Model failed to initialize. Server will continue running but enhancement won't work.")
    # We'll keep running but mark that we attempted initialization
    init_attempted = True

@app.route('/enhance', methods=['POST'])
def enhance_image():
    global upsampler
    
    # Check if upsampler is ready
    if upsampler is None:
        return jsonify({'error': 'Enhancement model is not initialized.'}), 500
    
    # Check if file was uploaded
    if 'file' not in request.files:
        return jsonify({'error': 'No file uploaded'}), 400
    
    try:
        # Read and validate image
        file = request.files['file']
        if file.filename == '':
            return jsonify({'error': 'Empty file submitted'}), 400
        
        # Process the image
        img = Image.open(file.stream).convert('RGB')
        img_array = np.array(img)
        
        # Check image size
        h, w = img_array.shape[:2]
        if h > 2000 or w > 2000:
            return jsonify({'error': 'Image too large. Maximum size is 2000x2000 pixels'}), 400
        
        # Enhance image
        output, _ = upsampler.enhance(
            img_array,
            outscale=4,  # 4x super-resolution
            alpha_upsampler='realesrgan'
        )
        
        # Convert to JPEG bytes
        img_byte_arr = io.BytesIO()
        Image.fromarray(output).save(img_byte_arr, format='JPEG', quality=95)
        img_byte_arr.seek(0)
        
        return send_file(img_byte_arr, mimetype='image/jpeg')
    
    except Exception as e:
        return jsonify({'error': f'Processing error: {str(e)}'}), 500

@app.route('/health', methods=['GET'])
def health_check():
    global upsampler, init_attempted
    
    # If we have the upsampler, we're ready
    if upsampler is not None:
        status = 'ready'
    # If we tried to initialize and failed, report failure
    elif init_attempted:
        status = 'failed'
    # Otherwise we're still in an unknown state
    else:
        status = 'initializing'
    
    status_info = {
        'status': status,
        'timestamp': time.time()
    }
    
    return jsonify(status_info)

@app.route('/')
def home():
    global upsampler
    return jsonify({
        'message': 'Image Enhancement API',
        'endpoints': {
            'POST /enhance': 'Process images (4x upscale)',
            'GET /health': 'Service status check'
        },
        'status': 'ready' if upsampler is not None else 'not ready'
    })

if __name__ == '__main__':
    # Start the Flask app
    app.run(host='0.0.0.0', port=5000)