Research overview: computer-vision individual identification from coat patterns

Identifying individual cats from their unique stripe, spot, or rosette patterns is the cornerstone of capture-recapture, and software increasingly assists this once wholly manual task. This overview surveys computer-vision tools such as Wild-ID, HotSpotter, and pattern-matching algorithms that compare new images against a catalogue to propose matches, using local feature descriptors and pattern geometry to rank candidates for human confirmation. It explains how these systems accelerate building individual capture histories for tigers, leopards, jaguars, ocelots, and other patterned felids, reduce matching errors and observer fatigue, and scale to large image sets. The overview also covers limitations, including sensitivity to image quality, pose, and lighting, and the move toward deep-learning re-identification. Reliable individual ID directly determines the validity of density estimates from camera traps.