摘要
近些年来,行人检测一直是计算机视觉领域内的一个研究热点。但在实际应用中,由于光照变化、复杂背景、行人遮挡等问题的存在,目前仍然缺乏一个鲁棒健壮的行人检测系统,可以满足实际应用中的需要。本文首先介绍了HOG特征和Adaboost算法,然后提出了一种基于HOG特征结合Adaboost算法的行人检测算法。最后在行人检测数据集上的实验结果表明本文提出的算法的检测准确率高达95%,并且在分辨率为640*480的视频上可以基本满足实时性的要求。
Pedestrian detection is a hot topic in the field of computer vision in recent years.But in practical applications,due to the illumination change,complex background,pedestrian occlusion and other problems,there is still a robust and robust pedestrian detection system,which can meet the needs of practical application.This paper first introduces the HOG features and Adaboost algorithm,and then proposes a pedestrian detection algorithm based on HOG features combined with Adaboost algorithm.Finally,the experimental results on pedestrian detection datasets show that the proposed algorithm achieves a detection accuracy of up to 95%,and can meet the real-time requirements in the video with resolution of 640*480.
作者
樊春年
杜卫平
刘艳荣
FAN Chun-nian;DU Wei-ping;LIU Yan-rong(Xinjiang Institute of Light Industry Technology,Urumqi 830021 China)
出处
《自动化技术与应用》
2018年第7期89-91,共3页
Techniques of Automation and Applications
