Dynamic fry counting based on multi-object tracking and one-stage detection

Hanyu Zhang, Weiran Li, Yanyu Qi, Haonan Liu, Zhenbo Li

Abstract: Highlights•A dynamic fry counting method based on multi-object tracking is studied.•YOLOv5-Nano can balance accuracy and complexity to realize fry detection.•The improved tracking algorithm can achieve 82.6 % tracking accuracy.•The method can be applied to GPU at 15–35 fps with 96.4 % counting accuracy.