单目视觉
计算机科学
正确性
单眼
计算机视觉
人工智能
机器视觉
姿势
算法
出处
期刊:Optik
[Elsevier BV]
日期:2017-02-01
卷期号:131: 72-78
被引量:12
标识
DOI:10.1016/j.ijleo.2016.11.075
摘要
In monocular vision system, non-coplanar target is extensively used for pose measurement. However, it remains a subject to be settled how to design and select the target parameters. This article proposes an optimized design method for non-coplanar target parameters to ensure that the monocular vision measuring system can attain optimal pose measurement accuracy. Firstly, a monocular-vision pose measurement model is established covering non-coplanar target parameters, and a pose measurement error model is deduced. On this basis, an objective function is put forward for optimizing the design of non-coplanar target parameters. Finally, simulations are conducted and the results verify the correctness of pose measurement error model and the effectiveness of the optimized design of target parameters.
科研通智能强力驱动
Strongly Powered by AbleSci AI