12,000-fps Multi-object Detection Using HOG Descriptor and SVM Classifier
Li Jianquan1,2; Yin Yingjie1,2; Liu Xilong1,2; Xu De1,2; Gu Qingyi1,2
2017-09
会议名称IEEE/RSJ International Conference on Intelligent Robots and Systems
会议日期September 24–28, 201
会议地点Vancouver, BC, Canada
摘要Abstract— This paper describes a high-frame-rate (HFR) vision system that can detect multiple objects in an image of 512×512 pixels at 12,000 frames per seconds (fps). An optimized algorithm is proposed based on conventional Histograms of Oriented Gradient  (HOG) descriptor and Support Vector Machine (SVM) classifier algorithms for hardware implementation. By implementing the proposed algorithm on a field-programmable gate array (FPGA) of a high-speed vision platform, multiobject in an image can be detected at 12,000 fps under complex background. In hardware implementation, 64 pixels were processed in parallel with 80 MHz camera clock. Source image and detection results can be transferred to personal computer (PC) in real-time for recording or post-processing. Our developed HFR multi-object detection system was verified by performing several evaluations.
文献类型会议论文
条目标识符http://ir.ia.ac.cn/handle/173211/20064
专题精密感知与控制研究中心_人工智能与机器学习
作者单位1.CASIA
2.UCAS
推荐引用方式
GB/T 7714
Li Jianquan,Yin Yingjie,Liu Xilong,et al. 12,000-fps Multi-object Detection Using HOG Descriptor and SVM Classifier[C],2017.
条目包含的文件 下载所有文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
12,000-fps Multi-obj(1422KB)会议论文 开放获取CC BY-NC-SA浏览 下载
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[Li Jianquan]的文章
[Yin Yingjie]的文章
[Liu Xilong]的文章
百度学术
百度学术中相似的文章
[Li Jianquan]的文章
[Yin Yingjie]的文章
[Liu Xilong]的文章
必应学术
必应学术中相似的文章
[Li Jianquan]的文章
[Yin Yingjie]的文章
[Liu Xilong]的文章
相关权益政策
暂无数据
收藏/分享
文件名: 12,000-fps Multi-object Detection Using HOG Descriptor and SVM Classifier.pdf
格式: Adobe PDF
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。