Duplicate Photo Scanner

The Duplicate Photo Scanner uses the dHash (difference hash) algorithm to create a compact 64-bit visual fingerprint for each image. First, the image is resized to 9×8 pixels and converted to grayscale, removing color information and size differences. Then the algorithm compares each pixel to its right neighbor — if the left pixel is brighter, it records a 1; otherwise, a 0. This produces a 64-bit binary hash that captures the image's gradient structure. Two images with similar visual content will produce similar hashes regardless of resolution, file format, slight cropping, or JPEG recompression. The Similarity Scanner uses a different approach for detailed side-by-side comparison of two specific images.

Photo deduplication is essential for managing large image libraries and freeing up storage space. Use this scanner to clean up your camera roll after a shoot, identify reposted or stolen images across collections, or find near-duplicate copies that differ only in resolution or compression. Photographers use it before archiving to remove burst-mode duplicates. Forensic investigators use it to find the same image distributed across multiple folders or devices. Content moderators use perceptual hashing to detect reposts. For verifying whether a specific image has been tampered with rather than copied, try the Authenticity Checker or ELA Scanner.

The sensitivity slider controls the Hamming distance threshold — the maximum number of differing bits between two 64-bit hashes for them to be considered duplicates. At sensitivity 1–3, only near-exact copies are detected (same photo resaved or slightly recompressed). At 5–10, the scanner catches resized versions, minor crops, and light exposure adjustments. At 15–20, even significantly edited versions of the same scene may be grouped together. Start at the default (10) and adjust based on your needs. For pixel-perfect comparison of two specific images, the Compare Photos tool provides overlay and slider views.

Perceptual hashing is just one layer of image analysis. Once you've identified duplicate groups, use the EXIF Checker to see which copy has the richest metadata — often the original. Check Privacy Score to ensure copies don't leak GPS coordinates or camera serial numbers. Run the File Hash Scanner on suspected duplicates — identical cryptographic hashes confirm byte-for-byte copies, while perceptual hashes catch visual copies that differ at the binary level. For batch metadata analysis across all your images, the Batch Scanner processes up to 20 files at once.