Python scripts that compares PCA mean vectors for altered images and unaltered images from raw files

pca-vectors.py

import cv2 as cv
import numpy as np
import rawpy


def perform_pca(filepath, alter_image=False):
    if alter_image:
        # Read the CR2 image with auto white balance or auto brightness
        rgb_image = rawpy.imread(filepath).postprocess()
    else:
        # Read the CR2 image without auto white balance or auto brightness
        rgb_image = rawpy.imread(file_path).postprocess(no_auto_bright=True, use_auto_wb=False, use_camera_wb=True)

    # Reshape the image data to a 2D array of RGB values
    x = rgb_image.reshape(-1, 3).astype(np.float32)

    # Perform PCA using OpenCV's PCACompute
    rgb_means, eigenvectors, eigenvalues = cv.PCACompute2(x, mean=None)

    # get the mean of each channel of the image using numpy
    rgb_means_numpy = x.mean(axis=0)

    return rgb_means, rgb_means_numpy


path_name = '/path/to/cr2-files/'  # Replace with your CR2 file path
filenames = ['IMG_1837.CR2', 'IMG_1839.CR2']  # Replace with your CR2 filenames

for filename in filenames:
    file_path = f'{path_name}{filename}'

    # Step 1: Perform PCA on the ALTERED CR2 image
    altered_rgb_means, altered_rgb_means_numpy = perform_pca(file_path, alter_image=True)

    # Step 2: Print the mean PCA values for the ALTERED CR2 image
    print(f"ALTERED image {filename}:")
    print(
        f"PCA mean vectors:  R: {altered_rgb_means[0][0]:.6f}   G: {altered_rgb_means[0][1]:.6f}   B: {altered_rgb_means[0][2]:.6f}")
    print(
        f" Numpy RGB means:  R: {altered_rgb_means_numpy[0]:.6f}   G: {altered_rgb_means_numpy[1]:.6f}   B: {altered_rgb_means_numpy[2]:.6f}")
    print()

    # Step 3: Perform PCA on the UNALTERED CR2 image
    unaltered_rgb_means, unaltered_rgb_means_numpy = perform_pca(file_path, alter_image=False)

    # Step 4: Print the mean PCA values for the UNALTERED CR2 image
    print(f"UNALTERED image {filename}:")
    print(
        f"PCA mean vectors:  R: {unaltered_rgb_means[0][0]:.6f}   G: {unaltered_rgb_means[0][1]:.6f}   B: {unaltered_rgb_means[0][2]:.6f}")
    print(
        f" Numpy RGB means:  R: {unaltered_rgb_means_numpy[0]:.6f}   G: {unaltered_rgb_means_numpy[1]:.6f}   B: {unaltered_rgb_means_numpy[2]:.6f}")

    print()
    print()

This is a Fake code! this is not PCA

This is a Fake code! this is not PCA

Fake comment. The code calls OpenCV’s PCACompute2.

Fake comment? I'm telling you this is not how PCA code is written. Get some basic education and ethics before dismissing feedback.

Fake comment? I'm telling you this is not how PCA code is written. Get some basic education and ethics before dismissing feedback.

Thanks for the feedback. I ran out and got some basic education and ethics. The education taught me this is how PCA code is written. The ethics taught me that harassing and defaming people is bad.

Get your trashy code reviewed by a decent capable person, if they're willing that is.
Creating photoshop images and calling them real based on a nice story exactly what defames a person, if they had any shame, they'd know better.