首页 | 本学科首页   官方微博 | 高级检索  
     

基于稳定区域梯度方向直方图的行人检测方法
引用本文:陈锐,彭启民.基于稳定区域梯度方向直方图的行人检测方法[J].计算机辅助设计与图形学学报,2012,24(3):372-377.
作者姓名:陈锐  彭启民
作者单位:1. 中国科学院软件研究所天基综合信息系统重点实验室 北京 100190;中国科学院研究生院 北京 100049
2. 中国科学院软件研究所天基综合信息系统重点实验室 北京 100190
摘    要:针对HOG算法采用简单均匀分布的块提取行人特征,导致过多冗余特征和检测效率较低问题,提出一种启发式的块生成算法.首先将待检测窗口划分成多个大小不一且交叉重叠的块,然后根据各个块的梯度方向直方图特征的稳定性对各个块进行排序,最后选取稳定性最高的多个块作为最终HOG算法的特征提取块.实验结果表明,该算法可取得较好的检测效果,并且检测速度得到了进一步的提高.

关 键 词:行人检测  梯度方向直方图  特征评价

Pedestrian Detection Based on HOG of Stable Area
Chen Rui , Peng Qimin.Pedestrian Detection Based on HOG of Stable Area[J].Journal of Computer-Aided Design & Computer Graphics,2012,24(3):372-377.
Authors:Chen Rui  Peng Qimin
Affiliation:1) 1)(Science and Technology on Integrated Information System Laboratory,Institute of Software,Chinese Academy of Sciences,Beijing 100190) 2)(Graduate University of Chinese Academy of Sciences,Beijing 100049)
Abstract:In this paper we propose a novel heuristic block-generating algorithm for reducing redundant features and increasing detection efficiency in pedestrian detection.Firstly,the detection window is divided into overlapped blocks with variant sizes.Then these blocks are sorted according to the stability of the histogram of oriented gradient for each block.Finally,HOG features from the most stable blocks can be extracted.Experimental results show that the proposed algorithm is faster and more accurate than the existing ones.
Keywords:pedestrian detection  histogram of oriented gradient  feature evaluation
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号