基于DeepSORT算法的肉牛多目标跟踪方法

Beef Cattle Multi-target Tracking Based on DeepSORT Algorithm

肉牛的运动行为反映其健康状况,在实际养殖环境下如何识别肉牛并对其进行跟踪,对感知肉牛的运动行为至关重要。基于YOLO v3改进算法(LSRCEM-YOLO),利用视频监控实现了实际养殖环境下的肉牛实时跟踪。该方法采用MobileNet v2作为目标检测骨干网络,根据肉牛分布不均、目标尺度变化较大的特点,提出通过添加长短距离语义增强模块(LSRCEM)进行多尺度融合,结合Mudeep重识别模型实现了肉牛多目标跟踪。结果表明:在目标检测方面,LSRCEM-YOLO的m AP值达到了92.3%,模型参数量仅为YOLO v3的10%,相比YOLO v3-tiny也降低了31.34%;在肉牛重识别方面,采用基于调整感受野的Mudeep模型,获得了更多的多尺度特征,其Rank-1指标达到了96.5%;多目标跟踪的多目标跟踪准确率相对于Deep SORT算法从32.3%提高到了45.2%,ID switch次数降低了69.2%。本文方法可为实际环境下的肉牛行为实时跟踪、行为感知提供技术参考。

The behavior of beef cattle reflects its health status. How to recognize and track beef cattle in a real breeding environment is very important to perceive the behavior of beef cattle. Wearable devices have limited accuracy in sensing motion behavior and are easily damaged,while monitoring devices are widely used in farms and have a long lifespan. Based on the improved YOLO v3 algorithm( LSRCEMYOLO),surveillance video was used to achieve real-time tracking of beef cattle in a real breeding environment. MobileNet v2 was used as the object detection backbone network. According to the uneven distribution of beef cattle and the large change of target scale,long-short range context enhancement module( LSRCEM) was proposed for multi-scale fusion,combined with the Mudeep ReID model to achieve multiple targets for beef cattle track. The experimental results showed that in beef cattle object detection,the m AP index of LSRCEM-YOLO reached 92. 3%,and the model parameter amount was only10% of YOLO v3,which was also reduced by 31. 34% compared with YOLO v3-tiny;in terms of beef cattle re-identification( Re ID),based on adjusting the Mudeep model of the receptive field obtained more multi-scale features,and its Rank-1 index reached 96. 5%. Compared with the original DeepSORT algorithm,the MOTA index of multi-target tracking was increased from 32. 3% to 45. 2%,and the number of ID switch was decreased by 69. 2%. This method can provide technical reference for real-time tracking and behavior perception of beef cattle in real environment.