Compare images and detect duplicates with perceptual hashing
Drag and drop 2 or more images. Or upload an entire folder of images for batch comparison.
Adjust the threshold (0-100%). Higher = only exact duplicates. Lower = includes visually similar images.
Click Scan. The tool computes perceptual hashes and compares them.
Duplicate/near-duplicate pairs are displayed with similarity score. Choose which to delete.
Find and remove duplicate photos from your collection (same photo saved multiple times, edited vs. original, etc.).
Clean up design asset folders by finding visually identical images with different names or formats.
Find unauthorized copies of your images that have been slightly modified (cropped, recompressed, color-adjusted).
Unlike cryptographic hashing (MD5, SHA), perceptual hashing generates similar hashes for visually similar images. Two images that are 90% similar will have similar pHash values.
Yes. Perceptual hash is resilient to resizing and minor cropping. But heavily cropped images may not match.
90%+ = near-exact duplicates. 70-90% = visually very similar (likely duplicates with minor edits). Below 70% = similar but possibly different images.