中国半导体行业协会封装分会会刊

中国电子学会电子制造与封装技术分会会刊

导航

电子与封装 ›› 2022, Vol. 22 ›› Issue (8): 080501 . doi: 10.16257/j.cnki.1681-1070.2022.0814

• 产品与应用 • 上一篇    下一篇

基于FPGA的双源无轨电车的改进型YOLO-V3模型*

董宜平1;谢达2;钮震3;彭湖湾2;贾尚杰2   

  1. 1. 中国电子科技集团公司第五十八研究所,江苏 无锡 214072;2. 无锡中微亿芯有限公司,江苏 无锡 214072;3. 凯博易控车辆科技苏州股份有限公司,江苏 苏州 215200
  • 收稿日期:2022-03-16 出版日期:2022-08-26 发布日期:2022-05-17
  • 作者简介:董宜平(1983—),男,江苏宜兴人,博士,高级工程师,现从事FPGA架构和应用、人工智能算法、片上网络等方面的研究。

Improved YOLO-V3 Model for Dual-Powered Trolley Bus Based on FPGA

DONG Yiping1, XIE Da2, NIU Zhen3, PENG Huwan2, JIA Shangjie2   

  1. 1. China Electronic Technology Group Corporation No. 58 Research Institute, Wuxi 214072, China; 2. East Technologies Inc., Ltd., Wuxi 214072, China; 3. eKontrol Vehicle Technology SuZhou Co., Ltd., Suzhou 215200, China
  • Received:2022-03-16 Online:2022-08-26 Published:2022-05-17

摘要: 为实现双源无轨电车对集电盒的智能识别和挂载,基于第三版传统黑暗网络的主干网络单次检测(YOLO-V3)网络模型,提出以轻量化移动网络为主干网络的改进型YOLO-V3网络。通过数据集的处理、模型的设计、训练环境的搭建等完成了网络的部署,然后对模型规模、识别精度和处理速度等指标进行比较。结果显示改进型YOLO-V3网络使用最小的计算资源得到最优精度。网络部署在FPGA内部中央处理器的分散处理单元中。实车测试结果表明,改进网络明显优于其他传统网络。

关键词: YOLO-V3网络, 移动网络, 目标检测, FPGA, 深度学习

Abstract: To realize the intelligent identification and mounting of the collector box for the dual-powered trolley bus, an improved you only look once - version 3 (YOLO-V3) model with MobileNet as main stem based on YOLO-V3 of traditional Darknet was introduced. Through dealing with data sets, modelling and establishing training environment, the network was deployed and compared with traditional methods about model scale, recognition accuracy and processing speed. The simulation results showed that the proposed YOLO-V3 network had the higher precision with lower overhead. The improved YOLO-V3 network was realized in data processing units on FPGA platform. The real trolley bus running test results showed that the improved network was better than other traditional networks.

Key words: YOLO-V3 network, MobileNet, object detection, FPGA, deep learning

中图分类号: