At present, the sorting of express parcels still requires manual participation in the operations of package supply, deviation correction and separation. In order to improve the automation of express parcel sorting, a multi-target tracking algorithm based on improved YOLOv5 and DeepSORT is proposed. Firstly, the SE attention mechanism is incorporated into the YOLOv5 network to enhance the model’s feature extraction capability for express parcels. Secondly, the EIOU loss function is employed as an alternative to the original CIOU loss function used in YOLOv5, aiming to improve the model’s convergence. Finally, the improved YOLOv5 model is utilized as the detector for DeepSORT to conduct multi-object tracking experiments on express parcels. Experimental results demonstrate that the improved YOLOv5 algorithm achieves a 4.5% improvement in mean average precision (mAP_0.5) compared to the original algorithm, along with a 1.3% increase in precision and a 6.3% increase in recall. When combined with DeepSORT, the proposed approach enabled accurate tracking of express parcels with good real-time performance.
Real-Time Detection and Tracking of Express Parcels Based on Improved YOLOv5+DeepSORT
Lect.Notes Computer
International Conference on Intelligent Robotics and Applications ; 2023 ; Hangzhou, China July 05, 2023 - July 07, 2023
2023-10-16
12 pages
Article/Chapter (Book)
Electronic Resource
English
Express parcels , YOLOv5 algorithm , DeepSORT algorithm Computer Science , Artificial Intelligence , Software Engineering/Programming and Operating Systems , Computer Applications , User Interfaces and Human Computer Interaction , Computer Communication Networks , Special Purpose and Application-Based Systems