Candid Camera: Using Visual Data as a New Way to Weigh Cattle By Vanessa Rotondo and Dr. Katie Wood, Department of Animal Biosciences, University of Guelph
T
here is an old saying in Hollywood: “the camera adds 10 pounds.” Although that may not be literally true, camera systems can in fact be used to predict body weight. A new research project at the University of Guelph has been developing the use of visual spectrum data from camera systems to automatically and accurately estimate body weight and growth in calves. Having accurate body weights on animals can be an important management tool for producers. Beyond benchmarking animal performance, animal weight can give insight into estimating nutrient requirements, be used as a health indicator, serve as a tool for genetic selection, and more. However, many farms either do not have access to scales and weighing infrastructure or do not have the time nor labour to regularly weigh their cattle. According to the 2016 Ontario Cow-Calf Survey, only 44 per cent of survey respondents owned a scale. This is where the base of this research starts. Is there another way to capture animal weights without the need for weighing infrastructure and labour associated with weighing cattle? The project brought together Dr. Medhat Moussa and research staff David Weales from the School of Engineering, Dr. Katie Wood and graduate student Vanessa Rotondo from the Department of
Private Treaty Black Angus Bulls and Bred Heifers Available
Animal Biosciences, and OMAFRA industry experts to develop a system that passively collects calf images to see if the images can be used to automatically identify key body points, which could be used to predict body weight. Although heart girth circumference has long been correlated with body weight and is the basis behind animal weigh tapes that many producers still use, Measuring over 19 body dimensions traits on young calves at the Ontario Beef Research there may be other Centre Measurements measurements that improve the accuracy of body weight prediction. To investigate the relevance of these other measurements, over 100 calves from the Ontario Beef Research Centre were weighed and 19 different body traits were measured weekly from two to eight weeks of age. Once measurements were collected, algorithms were generated to predict calf body weight. Measurements including heart girth, width over the pin bones, and width across the eyes, combined with other traits, were found to be able to predict calf weight with high accuracy.
Breeding for structural correctness, phenotype, growth, performance and calving ease. Pictures and videos available or a socially distanced farm visit can be arranged. (20 minutes southwest of Collingwood) Don & Betty McNalty and Family, Rob Roy Angus For further details call Don robroyangus@gmail.com Cell: 705-446-4740
18
O N T A R I O B E E F • M AY 2 0 2 1
Calf standing under the prototype camera system used to passively capture visual data which can be used to estimate body weight
