Monitoring System for Livestock Growth Based on Machine Vision Technology

Jing-Jing CHEN,Yan-Kun PENG
DOI: https://doi.org/10.13031/aim.20131586888
2012-01-01
Abstract:An imaging system for monitoring livestock growth based on machine vision technology was developed. Machine vision hardware-software system was self developed in the laboratory to predict the post-mortem meat yield of the pig. Two CCD cameras with Matrox cards were used to capture the image of live pig. The Matrox cards were controlled by self developed VC++ program. The images of the pig were then used to measure body measurements including body height (BH), body length (BL), chest width (CW) and hip width (HW). Based on image processing algorithm developed in Matlab and SPSS software, body weight (BW) and body measurements were used to predict the post mortem meat yield of the pig. A Graphical Users Interface was developed equipped with control system to acquire image of the pig. The results showed that body weight and body height can be used to estimate post-mortem meat yield with a high prediction accuracy. This research can be a milestone for development of machine vision technology to predict the meat yield from a live pig.
What problem does this paper attempt to address?