PDQ is a photo hashing algorithm that can turn photos into 256 bit signatures which can then be used to match other photos. The goal of this image hashing is to represent an image using a fixed-size representation (the hash) that preserves visual similarity, i.e., images that are visually similar should have similar hashes.
The PDQ hashes enable fuzzy deduplication of images and video (through thumbnails).
PDQ is a photo hashing algorithm that can turn photos into 256 bit signatures which can then be used to match other photos. The goal of this image hashing is to represent an image using a fixed-size representation (the hash) that preserves visual similarity, i.e., images that are visually similar should have similar hashes.
The PDQ hashes enable fuzzy deduplication of images and video (through thumbnails).