Automatic Retrieval of Specific Cows from Unlabeled Videos

TL;DR

This paper presents a system for automatic, hands-free identification and cataloging of individual cows in unstructured videos, combining a non-deep learning recognizer with a video processing pipeline.

Contribution

It introduces a novel system that enables cow identification in unlabeled videos without deep learning, facilitating automated herd monitoring.

Findings

Successfully identifies cows in unsegmented videos

Builds a comprehensive cow catalog from minimal input

Operates effectively in real-world dairy environments

Abstract

Few automated video systems are described in the open literature that enable hands-free cataloging and identification (ID) of cows in a dairy herd. In this work, we describe our system, composed of an AutoCattloger, which builds a Cattlog of dairy cows in a herd with a single input video clip per cow, an eidetic cow recognizer which uses no deep learning to ID cows, and a CowFinder, which IDs cows in a continuous stream of video. We demonstrate its value in finding individuals in unlabeled, unsegmented videos of cows walking unconstrained through the holding area of a milking parlor.