A pig behavior-tracking method based on a multi-channel high-efficiency attention mechanism

Given that the pig behavior reflects their health status, continuous and precise monitoring of behavior is important for effective health management and welfare protection. To mitigate potential tracking failures during analysis of video footage, we introduced a novel multi-target pig tracking method that consisted of detection and tracking components. The detection model was enhanced with an efficient attention mechanism and a Cross Stage Partial Darknet backbone network, which significantly improved detection accuracy. The tracking component used the Bytetrack algorithm to accurately track the movement trajectories of individual pigs. Together, these components were combined into the Dual-YOLOX-Tiny-ByteTrack (DYTB) model, which demonstrated superior performance in automatic monitoring of pig behaviors compared to previously published approaches. We established multi-object pig tracking datasets with 180,321 images to evaluate this method. The DYTB method achieved a pig detection accuracy of 98.3% and tracking accuracies of 95.3% and 97.1%. Compared to the YOLOX-Tiny-ByteTrack base model, DYTB showed a 3.4% improvement in multiple object tracking accuracy, making it a robust method for non-contact, intelligent monitoring of pig health and contributing to advances in precision livestock farming. • We first established an extensive multi-object pig tracking dataset with 180,321 images to address the limitations of existing datasets and enhance the generalization ability of the models. • We proposed a dual-branch high-efficiency channel attention-based convolutional module, referred to as Dual_ECA_BaseConv, resulting in Dual-YOLOX-Tiny with a pig detection accuracy of 98.3%. • We developed a novel multi-object pig detection and tracking algorithm by integrating the Dual-YOLOX-Tiny and Bytetrack algorithms, which demonstrated superior performance with a mAP value of 95.3% and an IDF1 score of 97.1% for pig tracking.