Improved military equipment identification algorithm based on YOLOv5 framework
Huibai Wang, Han Ji · 2023
In the field of military equipment target recognition, in order to solve the problems of small target detection, such as easy loss and slow detection speed, this paper proposes a target recognition and detection algorithm YOLO-M, which not only keeps high accuracy, but also reduces the number of parameters. Firstly, C3CMix module is used to replace part of the structure of head in YOLO v5, which reduces the number of parameters and effectively reduces the model size. Then, the detection accuracy is further improved by replacing the activation function in the structure. The experimental results show that the average accuracy of YOLO-M is 95.2%, the parameters are reduced by 18.8%, and the calculation is reduced by 14.5%. Therefore, the improved algorithm in this paper is more suitable for deployment in the field of military equipment target recognition.