微型电脑应用2026，Vol.42Issue(5)：277-280,4.
基于YOLOv5的目标轮廓运动特征检测方法
Feature Detection Method for Target Contour Motion Based on YOLOv5
侯相茹1
作者信息
- 1. 黑龙江外国语学院,信息工程学院,黑龙江,哈尔滨 150025
- 折叠
摘要
Abstract
The accurate recognition of target contour motion features is a critical technology in video analysis,particularly in fields of aerospace and manufacturing.Traditional feature detection methods often perform poorly in identifying and classifying targets,especially in multi-target environments and low-resolution challenges of small to medium-sized targets in videos.Therefore,this study proposes a target contour motion feature detection method based on YOLOv5.Utilizing YOLOv5,three-dimensional modeling of moving targets is achieved,and a staged approach is employed to extract target contour features.Ad-ditionally,convolutional neural networks are utilized to enhance feature learning,thus optimizing the detection process.The re-sults indicate that the proposed method achieves image frame detection at a speed of 29 frames per second with an average preci-sion of 91.74％,demonstrating its outstanding performance in detecting target contour motion features.关键词
目标轮廓/运动特征检测/YOLOv5/3D建模Key words
target contour/motion feature detection/YOLOv5/3D modeling分类
信息技术与安全科学引用本文复制引用
侯相茹..基于YOLOv5的目标轮廓运动特征检测方法[J].微型电脑应用,2026,42(5):277-280,4.
