Computer Vision and Deep Learningbased Framework for Cattle Monitoring

Precision livestock farming techniques enable us to obtain an accurate count of individual animals in a timely manner. Cattle counting and tracking are closely related to animal welfare in dairy farming. To accomplish this task, a computer vision framework is proposed in which ResNetV2 extracts features with the help of the optimizer YOLOv4, which significantly improves detection speed and accuracy.,The output will be updated in the centralised repository via IoT sensors and inconsistencies detected will be immediately communicated to the respective farmer. The proposed framework will assist government officials in taking corrective measures through livestock tracking in order to provide farmers with an efficient, quick, and accurate cattle count.