Accurate object recognition with assembling appearance and motion information

CORC > 光电技术研究所 > 中国科学院光电技术研究所 > 光电探测与信号处理研究室（五室）

	Accurate object recognition with assembling appearance and motion information
	Chang, Yongxin 1,2,3; Yu, Huapeng 1,2,3; Xu, Zhiyong 1; Zhang, Jing 2; Gao, Chunming 2
刊名	Mathematical Problems in Engineering
	2014
卷号	2014 页码:195941
ISSN号	1024123X
通讯作者	Chang, Yongxin
中文摘要	How to effectively detect object and accurately give out its visible parts is a major challenge for object detection. In this paper we propose an explicit occlusion model through integrating appearance and motion information. The model combines together two parts: part-level object detection with single frame and object occlusion estimation with continuous frames. It breaks through the performance bottleneck caused by lack of information and effectively improves object detection rate under severe occlusion. Through reevaluating the semantic parts, the detecting performance of partial object detectors is largely enhanced. The explicit model enables the partial detectors to have the capability of occlusion estimation. By discarding the geometric representation in rigid single-angle perspective and applying effective pattern of objective shape, our proposed approaches greatly improve the performance and robustness of similarity measurement. For validating the performance of proposed methods, we designed a comparative experiment on challenging pedestrian frame sequences database. The experimental results on challenging pedestrian frame sequence demonstrate that, compared to the traditional algorithms, the methods proposed in this paper have significantly improved the detection rate for severe occlusion. Furthermore, it also can achieve better localization of semantic parts and estimation of occluding.
英文摘要	How to effectively detect object and accurately give out its visible parts is a major challenge for object detection. In this paper we propose an explicit occlusion model through integrating appearance and motion information. The model combines together two parts: part-level object detection with single frame and object occlusion estimation with continuous frames. It breaks through the performance bottleneck caused by lack of information and effectively improves object detection rate under severe occlusion. Through reevaluating the semantic parts, the detecting performance of partial object detectors is largely enhanced. The explicit model enables the partial detectors to have the capability of occlusion estimation. By discarding the geometric representation in rigid single-angle perspective and applying effective pattern of objective shape, our proposed approaches greatly improve the performance and robustness of similarity measurement. For validating the performance of proposed methods, we designed a comparative experiment on challenging pedestrian frame sequences database. The experimental results on challenging pedestrian frame sequence demonstrate that, compared to the traditional algorithms, the methods proposed in this paper have significantly improved the detection rate for severe occlusion. Furthermore, it also can achieve better localization of semantic parts and estimation of occluding.
学科主题	Semantics
收录类别	SCI ; EI
语种	英语
WOS记录号	WOS:000345040300001
内容类型	期刊论文
源URL	[http://ir.ioe.ac.cn/handle/181551/5074]
专题	光电技术研究所_光电探测与信号处理研究室（五室）
作者单位	1.Institute of Optics and Electronics, Chinese Academy of Sciences, Chengdu, China 2.School of Optoelectronic Information, University of Electronic Science and Technology of China, Chengdu, China 3.Graduate University, Chinese Academy of Sciences, Beijing, China
推荐引用方式 GB/T 7714	Chang, Yongxin,Yu, Huapeng,Xu, Zhiyong,et al. Accurate object recognition with assembling appearance and motion information[J]. Mathematical Problems in Engineering,2014,2014:195941.
APA	Chang, Yongxin,Yu, Huapeng,Xu, Zhiyong,Zhang, Jing,&Gao, Chunming.(2014).Accurate object recognition with assembling appearance and motion information.Mathematical Problems in Engineering,2014,195941.
MLA	Chang, Yongxin,et al."Accurate object recognition with assembling appearance and motion information".Mathematical Problems in Engineering 2014(2014):195941.