Perceptual Image Hash API
Fingerprint images for near-duplicate detection and similarity. Compute the three classic perceptual hashes — aHash (average), dHash (difference) and pHash (DCT-based) — as 64-bit hex values for any image (by URL or base64), then compare two images to get the Hamming distance and a 0-100 similarity score per algorithm, with a likely-same flag. Unlike a cryptographic hash, perceptual hashes stay close when images are resized, recompressed or lightly edited — so you can spot duplicates, find re-uploads, cluster similar pictures and power reverse-image matching. Fully local (no third-party service), nothing stored. Supports PNG, JPEG, BMP, TIFF and GIF. Live. 3 endpoints. Distinct from basic image-info/resize and from string-similarity tools.
api.oanor.com/imghash-api
