PixelPeep - Intelligent Image Comparison and find original one from group of images

[joseph-mv]

This PR introduces the complete initial version of PixelPeep, a tool designed to intelligently compare and score images based on a histogram and imagehash. PixelPeep intelligently groups similar images and attempts to identify the original image from each group using a combination of histogram comparison and imagehash.

Features

Image Preprocessing

• Upload and decode images with UploadFile

Perceptual Similarity & Hashing

• Computes perceptual hash (pHash) for robust similarity comparison

• Calculates Hamming distance between image hashes to group visually similar images

Scoring System for Originality

• Normalizes all metrics safely (with zero-sum checks)
• Uses a weighted scoring algorithm to rank images within each group

Output

• Groups similar images based on threshold value

• For each group, identifies the most likely original image using the scoring system

Technologies Used:

• Python
• React
• Numpy,pillow,imageHash,openCV