Researchers at the University of Arkansas have developed an innovative AI tool named CattleFever, designed to estimate the body temperature of cattle using thermal images. This advancement allows ranchers to efficiently monitor herd health and enhance animal welfare, reducing the stress associated with traditional rectal temperature measurement. The tool leverages a unique dataset, CattleFace-RGBT, created from thousands of thermal videos and images of calves, pinpointing temperature readings from facial landmarks like the eyes and nostrils. By employing advanced machine learning, specifically random forest regression, CattleFever accurately predicts temperatures within one degree of traditional thermometers. While the initial demonstrations were conducted with calves facing the camera, researchers aim to adapt the system for real-world conditions where cattle are in motion. By sharing their dataset with the research community, they encourage further innovations in livestock health monitoring. This groundbreaking work marks a significant step toward automated herd health management.
Source link
Share
Read more