app.py

import numpy as np
import gradio as gr
from skimage import io, color
from numpy.linalg import norm
from PIL import Image

def svd_compress(image, k):
    """Compress the image using SVD by keeping only the top k singular values."""
    U, S, Vt = np.linalg.svd(image, full_matrices=False)
    compressed_image = np.dot(U[:, :k], np.dot(np.diag(S[:k]), Vt[:k, :]))
    return compressed_image

def process_image(image, k):
    """Process the uploaded image, compress it using SVD, and return the result."""
    # Convert PIL Image to NumPy array
    image_np = np.array(image)
    
    # Convert to grayscale
    gray_image = color.rgb2gray(image_np)
    
    # Compress the image
    compressed_image = svd_compress(gray_image, k)
    
    # Convert compressed image back to PIL Image for Gradio output
    compressed_image_pil = Image.fromarray((compressed_image * 255).astype(np.uint8))
    
    return compressed_image_pil

# Gradio interface
gr_interface = gr.Interface(fn=process_image,
                            inputs=[gr.inputs.Image(type="pil"), gr.inputs.Slider(1, 100, step=1, default=50)],
                            outputs="image",
                            title="Interactive Image Compression using SVD",
                            description="Upload an image and adjust the compression rank to see the compressed version.")

gr_interface.launch()