Dataset_creater.py

# -*- coding: utf-8 -*-
"""Untitled6.ipynb

Automatically generated by Colab.

Original file is located at
    https://colab.research.google.com/drive/1c5MoZkw2FXiG_RcYIe-StsDDLmMKYDqw
"""

import os
import cv2
import numpy as np
from PIL import Image
from tqdm import tqdm

# Function to apply Gaussian blur to an image
def apply_gaussian_blur(image, kernel_size=(5, 5), sigma=1):
    """Applies Gaussian blur to the input image.

    Parameters:
    - image: Input image (numpy array).
    - kernel_size: Size of the Gaussian kernel.
    - sigma: Standard deviation for the Gaussian kernel.

    Returns:
    - Blurred image.
    """
    return cv2.GaussianBlur(image, kernel_size, sigma)

# Create directories for saving the paired images
def create_dirs(output_dir, sharp_dir_name='sharp', blurred_dir_name='blurred'):
    """Creates directories for storing sharp and blurred images."""
    sharp_dir = os.path.join(output_dir, sharp_dir_name)
    blurred_dir = os.path.join(output_dir, blurred_dir_name)

    os.makedirs(sharp_dir, exist_ok=True)
    os.makedirs(blurred_dir, exist_ok=True)

    return sharp_dir, blurred_dir

# Process and save the paired dataset
def create_paired_dataset(input_dir, output_dir, kernel_size=(15, 15), sigma=3):
    """Creates a paired dataset of sharp and blurred images.

    Parameters:
    - input_dir: Directory containing the original (sharp) images.
    - output_dir: Directory where the sharp and blurred images will be saved.
    - kernel_size: Size of the Gaussian kernel.
    - sigma: Standard deviation for the Gaussian kernel.
    """
    sharp_dir, blurred_dir = create_dirs(output_dir)

    # Loop through all images in the input directory
    for img_name in tqdm(os.listdir(input_dir)):
        img_path = os.path.join(input_dir, img_name)

        # Open the image and convert it to a numpy array
        image = np.array(Image.open(img_path))

        # Apply Gaussian blur to the image
        blurred_image = apply_gaussian_blur(image, kernel_size, sigma)

        # Convert back to PIL Image for saving
        sharp_image_pil = Image.fromarray(image)
        blurred_image_pil = Image.fromarray(blurred_image)

        # Save the sharp and blurred images in their respective directories
        sharp_image_pil.save(os.path.join(sharp_dir, img_name))
        blurred_image_pil.save(os.path.join(blurred_dir, img_name))

# Example usage
input_dir = '/content/drive/MyDrive/Faces'  # Directory containing your sharp images
output_dir = '/content/'  # Directory to save the sharp and blurred images

# Create paired dataset with Gaussian blur (kernel size 5x5 and sigma 1)
create_paired_dataset(input_dir, output_dir, kernel_size=(25, 25), sigma=3)

import shutil

# Path to the folder you want to zip
folder_to_zip = '/content/sharp'

# Output path for the zip file (without extension)
output_zip = '/content/sharp_zip'

# Create a zip archive
shutil.make_archive(output_zip, 'zip', folder_to_zip)

print(f"Folder zipped as {output_zip}.zip")