ABSTRACT
Electronic systems for automatically weighing pigs within their pens have been developed for modern production. These modern equipments offer accurate and continual measurement of pig weights. Modern computer and cameras offer accurate and convenient methods for measuring dimensions, which is related to the weights. We can use computer vision technologies to monitor performance of growing pigs by calculating their weights from areas and dimensions measured from photographic images. This paper presents a method that estimates pig weights from images obtained by two cameras, one on the top of the ceiling and one on a sidewall. We calculated the projected areas and heights of pigs through computer vision analysis, and then found a close correlation between the weights and these numbers. The results show that pig weights can be calculated accurately from the projected areas and the heights. From experiments we obtained a regressive formulation, which is used in calculation of the pig weights.
