主办:陕西省汽车工程学会
ISSN 1671-7988  CN 61-1394/TH
创刊:1976年

Automobile Applied Technology ›› 2024, Vol. 49 ›› Issue (6): 78-83.DOI: 10.16638/j.cnki.1671-7988.2024.006.014

• Design and Research • Previous Articles    

Optimization of Vehicle Tire Fastener Detection Method Based on Improved YOLO_v3 Algorithm

ZHANG Baoyu   

  1. School of Intelligent Manufacturing, Jiangsu Food & Pharmaceutical Science College
  • Published:2024-03-27
  • Contact: ZHANG Baoyu

基于改进 YOLO_v3 算法的车辆轮胎紧固件 检测方法优化

张宝玉   

  1. 江苏食品药品职业技术学院 智能制造学院
  • 通讯作者: 张宝玉
  • 作者简介:张宝玉(1987-),男,硕士,讲师,研究方向为传感器与智能算法,E-mail:879727881@qq.com。
  • 基金资助:
    江苏省高等学校自然科学研究面上项目(20KJD510007)。

Abstract: Tire disassembly is one of the necessary steps for repairing tires in the auto repair industry. Due to the excessive heavy truck tires, the lumbar spine is damaged by the lumbar spine. The EfficientNet-B4 lightweight network structure algorithm is now replaced with the main part of the YOLO_v3 algorithm (DarkNet-53) to realize the identification of tire fasteners in various models. The replaced network parameters are largely reduced, and the fastener sample training speed is accelerated. YOLO_v3 uses a duplex cross-entropy loss function to classify and calculate the loss of positive and negative samples. The Focal loss classification loss function is now replaced by the binary cross-entropy loss function, so as to propose a new neural network model (YOLO_v3-Nut). The experimental results show that the YOLO_v3-Nut model is better than the YOLO_v3 model in terms of training and recognition speed characteristics, and the model structure of this article is reduced by 43.01% compared with the YOLO_v3 model storage space. The speed is 36 fps, which is enough to complete the identification of tire fasteners in various models.

Key words: Tire fastener; YOLO_v3; EfficientNet; Automatic disassembly; Categorical loss function

摘要: 轮胎拆装是汽修行业修补轮胎必要步骤之一,由于重型货车轮胎过重导致工人腰椎受 到严重损伤。现将 EfficientNet-B4 轻量型网络结构算法替换 YOLO_v3 算法的主干部分 (DarkNet-53),从而实现各类车型轮胎紧固件的识别。替换后的网络参数大量减少,紧固件 样本训练速度加快。YOLO_v3 使用二分类交叉熵损失函数对正负样本分类并计算损失,现使 用 Focal loss 分类损失函数替换二分类交叉熵损失函数,从而提出一种新的神经网络模型 (YOLO_v3-Nut)。实验结果表明,YOLO_v3-Nut 模型在训练与识别速度特性上都更优于 YOLO_v3 模型,且文中的模型结构比 YOLO_v3 模型储存空间减少了 43.01%,算法平均准确 率(MAP)为 93.2%,同时检测速度为 36 fps,足够完成各类车型轮胎紧固件的识别。

关键词: 轮胎紧固件;YOLO_v3;EfficientNet;自动拆卸;分类损失函数