StrongSort (OSNET) cooperates with YOLOv5, v7, v8 to achieve multi-target tracking detailed explanation
- This project contains a highly configurable two-stage tracker that can be adjusted for different deployment scenarios. Detections generated by YOLOv5, YOLOv7, YOLOv8 (a series of object detection architectures and models pre-trained on the COCO dataset) are passed to StrongSORT, which combines motion and appearance information based on OSNet to track objects. It can track any object that your Yolov5 model is trained to detect.
- In this article
yolov7
the tracking has not been run through, butyolov5、yolov8
the tracking and counting functions are directly available - Among them, the yolov5 tracking results are shown in the figure below. For other results, please see the article for details.
strongnet + yolov5 achieves multi-target tracking
Environment setup
#新建虚拟环境
conda create -n v7_sort python=3.9