《计算机应用》唯一官方网站 ›› 2023, Vol. 43 ›› Issue (1): 98-103.DOI: 10.11772/j.issn.1001-9081.2021101857

• 人工智能 • 上一篇    下一篇

基于嵌入式Jetson TX2的高原鼠兔目标检测

陈海燕, 贾明明, 赵文力, 王婵飞   

  1. 兰州理工大学 计算机与通信学院,兰州 730050
  • 收稿日期:2021-11-02 修回日期:2021-12-20 发布日期:2021-12-31
  • 作者简介:陈海燕(1978—),女,甘肃陇西人,副教授,博士,主要研究方向:模式识别、人工智能、图像处理 email:chenhaiyan@sina.com;贾明明(1994—),男,甘肃白银人,硕士研究生,主要研究方向:模式识别、人工智能;赵文力(1996—),男,甘肃天水人,硕士研究生,主要研究方向:模式识别、人工智能;王婵飞(1984—),女,甘肃会宁人,副教授,博士,主要研究方向:信号检测与估计;
  • 基金资助:
    国家自然科学基金资助项目(62161019, 62061024)。

Target detection of Ochotona curzoniae based on embedded Jetson TX2

CHEN Haiyan, JIA Mingming, ZHAO Wenli, WANG Chanfei   

  1. School of Computer and Communication, Lanzhou University of Technology, Lanzhou Gansu 730050, China
  • Received:2021-11-02 Revised:2021-12-20 Online:2021-12-31
  • Contact: CHEN Haiyan, born in 1978, Ph. D., associate professor. Her research interests include pattern recognition, artificial intelligence, image processing
  • About author:JIA Mingming, born in 1994, M. S. candidate. His research interests include pattern recognition, artificial intelligence;ZHAO Wenli, born in 1996, M. S. candidate. His research interests include pattern recognition, artificial intelligence;WANG Chanfei, born in 1984, Ph. D., associate professor. Her research interests include signal detection and estimation;
  • Supported by:
    This work is partially supported by National Natural Science Foundation of China (62161019, 62061024).

摘要: 高原鼠兔目标检测是对其进行种群数量统计及种群动态变化研究的基础,但传统的高原鼠兔智能监测系统的目标检测硬件设备大,在抽样采集数据时移动性较弱。针对此问题,提出一种可部署到便携式设备Jetson TX2上的基于改进YOLOv3模型的目标检测方法。该方法将YOLOv3的主干网络DarkNet53替换成MobileNet,并利用剪枝、微调等方法构建轻量级高原鼠兔目标检测模型,再将轻量化模型部署到Jetson TX2上。自然场景下高原鼠兔目标检测实验的结果表明:所提方法的检测平均精度(AP)、每秒检测帧数(FPS)和模型大小分别为97.36%、36和14.88 MB,优于主干网络替换后未裁剪的YOLOv3模型及原始YOLOv3模型,相较于原YOLOv3模型,AP在仅下降1.05个百分点的情况下,FPS提升了620%,模型大小压缩了93.67%,能够部署在便携设备上进行实时且准确的高原鼠兔目标检测。

关键词: 目标检测, YOLOv3, 轻量化, 模型剪枝, Jetson TX2, 高原鼠兔

Abstract: Target detection of Ochotona curzoniae is the basis for its population statistics and population dynamic changes research, but the traditional intelligent monitoring systems of Ochotona curzoniae has a large target detection hardware equipment and weak mobility in sampling and collecting data. To solve the above problems, an improved YOLOv3? based target detection method that can be deployed to the portable device Jetson TX2 was proposed. The lightweight Ochotona curzoniae target detection model was constructed by replacing Darknet53, backbone network of YOLOv3, with MobileNet and using pruning and fine?tuning methods to further reduce the model size. Next, the model was deployed on the portable target detection device Jetson TX2. Experimental results of Ochotona curzoniae target detection in natural scenes show that the proposed method has the detection Average Precision (AP), detection Frames per Second (FPS) and model size of 97.36%, 36 and 14.88 MB, respectively, which are better than those of the improved YOLOv3 model without pruning and the original YOLOv3 model; with the AP only reduced by 1.05 percentage points, the proposed method has the detection speed improved by 620% and the model size compressed by 93.67%, which can be deployed to portable devices for real-time and accurate detection of Ochotona curzoniae.

Key words: target detection, YOLOv3, lightweight, model pruning, Jetson TX2, Ochotona curzoniae

中图分类号: