verify_data.py

from PIL import Image
import os
from config import DATA_PATH

def verify_dataset_images():
    print(f"Verifying images in dataset directory: {DATA_PATH}")
    if not os.path.exists(DATA_PATH):
        print(f"Error: Dataset path '{DATA_PATH}' does not exist.")
        return

    corrupted_count = 0
    total_count = 0

    for person_name in os.listdir(DATA_PATH):
        person_dir = os.path.join(DATA_PATH, person_name)
        if not os.path.isdir(person_dir):
            continue
            
        print(f"Checking images for subject: {person_name}")
        
        for img_name in os.listdir(person_dir):
            if img_name.lower().endswith((".jpg", ".png", ".jpeg")):
                total_count += 1
                img_path = os.path.join(person_dir, img_name)
                
                try:
                    # Check if the image is valid
                    with Image.open(img_path) as img:
                        img.verify()
                        
                    # Also try converting PNGs to standard saving format if needed
                    if img_name.endswith(".png"):
                        with Image.open(img_path) as img_save:
                            img_save.save(img_path, "PNG")

                except Exception as e:
                    print(f"-> {img_name}: Corrupted or wrong format ({e})")
                    corrupted_count += 1

    print("-" * 30)
    print(f"Verification Complete. Total images: {total_count}")
    print(f"Corrupted images found: {corrupted_count}")

if __name__ == "__main__":
    verify_dataset_images()