摘要
首先引入主动视觉空间的概念 ,即用系统本身运动自由度来表达摄像机焦点到空间点的视线 .其次 ,推导了把任意摄像机图像点变换到主动视觉空间的算法 .最后 ,采用三角形测量原理对图像特征点进行匹配和定位 .因为系统中的各个摄像机独立注视目标 ,摄像机可以采用较小的视野 ,这为提高定位精度提供了可能性 .
Firstly, a terminology called active visual space, which represents feature points by the motion freedoms of an active visual system, is introduced. Secondly, we derive an algorithm, by which all the image points in the image space of a single active camera are translated into active visual coordinates. Finally, a rule for feature point matching and positioning is proposed on the analysis of stereo triangulation. Because the cameras in this system gaze object independently, just small scene is required. This feature provides the possibility to improve the positioning precision. Numerical simulation is presented to verify the scheme.
出处
《东南大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2002年第1期59-63,共5页
Journal of Southeast University：Natural Science Edition
基金
国家计委 2 11工程资助项目 (2 10 2 0 0 2 0 0 1)
