Automatic detection of mounting behaviours among pigs using image analysis

摘要: Automatic mounting events detection among pigs is presented.Ellipse fitting algorithms were used to localize each pig in the image.The Euclidean distances between head, tail and sides of obtained.Major minor axis lengths altered during events. Excessive behaviours amongst cause a high risk poor welfare, arising from skin lesions, lameness stress, economic losses reduced performance. The aim this study was develop method for automatic under commercial farm conditions by means image processing. Two pens selected monitored 20days top view cameras. recorded video then visually analysed selecting behaviours, extracted images files subsequently An ellipse technique applied image. intersection points major fitted shape defining pig. head tail, sides, length utilized development an algorithm automatically identify event. proposed could detect with level sensitivity, specificity accuracy, 94.5%, 88.6% 92.7%, respectively. results show that it possible use machine vision techniques order conditions.