中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
6D Pose Estimation Based on Multiple Appearance Features from Single Color Image

文献类型:会议论文

作者Zhu F(朱枫); Pan W(潘旺); Zhang LM(张丽敏); Hao YM(郝颖明)
出版日期2017
会议名称7th Annual IEEE International Conference on Cyber Technology in Automation, Control, and Intelligent Systems (IEEE-CYBER 2017)
会议日期July 31 - August 4, 2017
会议地点Hawaii, USA
页码406-411
通讯作者Pan W(潘旺)
中文摘要Template matching methods are widely used to recognize instances of an untextured object and determine the poses. But in many practical cases the false positive rate is high because of similar interference characteristic in images especially in complex and clutter scene. A method based on multiple appearance features is presented in the paper to recognize poorly textured 3D objects and estimate their 6D pose in single color image. The object is distinguished from cluttered environment by multi-features which include the color, size and aspect ratio. Image pyramid is introduced here to improve efficiency. And the CAD model of the object is used to generate a hierarchical shape model in the offline phase for estimating the 6D pose accurately. Lastly the approach is evaluated in our application and publicly available dataset, and achieves good performance both in speed and recognition rate compared with state-of-the-art methods.
产权排序1
会议录7th Annual IEEE International Conference on Cyber Technology in Automation, Control, and Intelligent Systems (IEEE-CYBER 2017)
会议录出版者IEEE
会议录出版地New York
语种英语
ISBN号978-1-5386-0489-2
源URL[http://ir.sia.cn/handle/173321/21303]  
专题沈阳自动化研究所_光电信息技术研究室
作者单位1.Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang 110016, China
2.The Key Lab of Image Understanding and Computer Vision, Liaoning Province, Shenyang 110016, China
3.Key Laboratory of Opto-Electronic Information Processing, CAS, Shenyang 110016, China
4.University of Chinese Academy of Sciences, Beijing 100049, China
推荐引用方式
GB/T 7714
Zhu F,Pan W,Zhang LM,et al. 6D Pose Estimation Based on Multiple Appearance Features from Single Color Image[C]. 见:7th Annual IEEE International Conference on Cyber Technology in Automation, Control, and Intelligent Systems (IEEE-CYBER 2017). Hawaii, USA. July 31 - August 4, 2017.

入库方式: OAI收割

来源:沈阳自动化研究所

浏览0
下载0
收藏0
其他版本

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。