基于SURF算法的目标跟踪应用技术研究
Target Tracking Application Technology Research Based On SURF Algorithm
王禹程 1王玉柱 2杨鲤源1
作者信息
- 1. 南京电子工程研究所,南京,210023
- 2. 北京市国家保密局,北京,100031
- 折叠
摘要
智慧城市是当前的一个热点话题,它的实现离不开视频监控的智能化,视频的图像目标匹配是智能视频监控中的重要内容.基于加速鲁棒性特征(SURF)局部特征的算法具有很好的稳定性,本文重点研究基于SURF算法的目标跟踪算法.在标清监控视频中,由于目标尺寸较小,使用SURF特征难以提取到足够的特征点,导致目标跟踪无法实现.本文通过对SURF算法进行改进以增加获取的SURF特征数,且改进的SURF算法在不同光线、旋转、尺度、视角变化条件下都能提取到比SURF算法更多的特征点,从而实现较好的跟踪效果.
Abstract
Smart city is a hot topic nowadays,but the realization of smart city cannot be separated from the intelligence of video surveillance.The image target matching is an important part of intelligent video sur-veillance.Due to the excellent stability of the algorithm based on Speeded-Up Robust Features,this paper will focus on theresearch of target tracking algorithm based on SURF algorithm.In standard definition surveillance video,due to the small size of the target,it is difficult to extract enough feature points using SURF features,resulting in the failure of target tracking.This paper improves the SURF features obtained,and the improved SURF algorithm can extract more feature points than the SURF algorithm under different light changes,rota-tion changes,scale changes and perspective changes,thus achieving better tracking effect.
关键词
SURF算法/特征提取/目标跟踪Key words
SURF algorithm/feature extraction/target tracking引用本文复制引用
出版年
2024
NETL
NSTL
万方数据