基于定向边界框标注的猪只多目标跟踪方法

打开文本图片集

中图分类号：TP391.4；S828 文献标志码：A 文章编号：1001-411X（2025）06-0821-11

Abstract： 【Objective】 To address the issue of pig trajectory loss caused by prolonged severe occlusion， we propose a multi-object tracking method for pigs based on oriented bounding box （OBB） annotation to enhance the tracking performance of pigs in ocluded scenarios. 【Method】 First， the YOLOv11n baseline model was improved by integrating a C3k2_DualConv convolutional network and a bidirectional feature pyramid network （BiFPN）， thereby constructing a pig detection model named YOLO-DB. Second， building upon the BoT-SORT tracking algorithm， an enhanced matching strategy was developed using a trajectory frame analysis mechanism， incorporating posture consistency features beforeand after occusion，and adopting a probabilistic intersection over union （ProbIoU）-based target matching mechanism optimized for OBB annotation. Finally， the YOLO-DB algorithm and improved BoT-SORT tracking algorithm were integrated to recover lost trajectories. 【Result】 The YOLO-DB algorithm achieved precision， recall and mAP50 of 96.5% ， 95.6% and 97.3% 0 respectively，representing increases of2.7，1.2and 1.4 percentage points compared to the baseline model， while its parameter count was reduced by 2.4% .The improved trackingalgorithm attained scores of 82.4% for higherorder tracking accuracy （HOTA）， 97.7% for multiple object tracking accuracy （MOTA）， and 89.1% for identification F1 score （IDF1），outperforming the baseline model by O.9，1.3，and 5.4 percentage points， respectively. Additionally， identity switches （IDS）， false positives （FP），and false negatives （FN） were significantly reduced. 【Conclusion】The proposed algorithm efectively resolves the problem of pig trajectory loss caused by prolonged occusion and significantly enhances tracking performance in ocluded scenarios. It provides an eficient and reliable technical solution for intelligent management in large-scale pig farms.

Key words： Multi-object tracking; Oriented bounding box; YOLOv1ln; Pig; BoT-SORT

随着计算机视觉和深度学习技术的快速发展，多目标跟踪技术在规模猪场中的应用日益广泛。（剩余15362字）