首页 | 杂志简介 | 征稿简则 | 投稿指南 | 常见问题 | 名词解释 | 刊物订阅 | 联系我们 | English
基于聚合通道特征的红外行人检测方法
投稿时间:2018-03-23  修订日期:2018-03-28  点此下载全文
引用本文:石永彪,张湧.基于聚合通道特征的红外行人检测方法[J].红外,2018,39(5):42~48
摘要点击次数: 33
全文下载次数: 47
作者单位E-mail
石永彪 中国科学院上海技术物理研究所 hainu2008@163.com 
张湧 中国科学院上海技术物理研究所  
中文摘要:基于红外图像的行人检测技术在夜间场景监控、汽车夜间辅助驾驶等相关领域具有重要的作用,然而受红外图像分辨率低、信噪比高等因素影响,当前的很多方法性能不佳。提出了一种基于图像特征通道的红外行人检测算法。利用快速特征金字塔技术在红外图像上进行了滑动窗口检测。实验结果证明,相对于其他常规算法,该算法在实时性和鲁棒性上都有很大的提升。
中文关键词:红外图像  行人检测  快速特征金字塔
 
Algorithm for Infrared Pedestrian Detection Based on Aggregated Channel Features
Abstract:The pedestrian detection technology based on infrared images has played an important role in night surveillance, night-time driver assistance and other areas. However, because of low resolution and high signal-to-noise ratio, most of the existing infrared pedestrian detection methods have poor performance. An alternative infrared pedestrian detection algorithm based on image feature channels is proposed. In the method, fast feature pyramids are used to implement sliding window detection in infrared images. The results show that compared with other conventional algorithms, this algorithm is improved greatly in real-time performance and robustness.
keywords:infrared image  pedestrian detection  fast feature pyramids
查看全文  查看/发表评论  下载PDF阅读器

版权所有:《红外》编辑部

北京勤云科技发展有限公司