一种基于一维和二维相机的混合光学定位方法  

A Combined Optical Positioning Method Based on 1-D and 2-D Cameras

作　　者：吴剑[1] 李昂[1] 

机构地区：[1]清华大学深圳研究生院生物医学工程研究所,深圳518055

出　　处：《生物医学工程学杂志》2015年第6期1310-1316,共7页Journal of Biomedical Engineering

基　　金：深圳市知识创新计划基础研究资助项目(JCY20130402145002404)

摘　　要：光学空间定位系统在诸多领域都有着广泛应用。本文提出了一种结合使用线阵相机和面阵相机的方法来进行空间定位。线阵相机的一维数据可以快速处理,提取出目标点的坐标。利用这个坐标来辅助面阵相机上的数据处理,减少面阵相机上目标的搜索量和计算量,同时还可以对多目标点定位的匹配过程提供帮助,进而达到加快计算速度的目的。通过本方法可以达到1.608mm的定位精度,约为测量范围的千分之一;而且可以同时对多目标点进行定位,即使在目标点之间并没有几何约束的条件下,该方法对4目标点的定位时间也可达23.87 ms,即为41.9fps。Optical positioning system has a wide range of application in many fields. This paper presents an optical positioning method which combines linear and plane cameras. 1-D image signal of linear cmm ra is used to get 1-D co- ordinates of the targets, which can be processed quickly. We can use it to assist processing the 2-D image signals of plane camera under some constraints. Linear camera can reduce the amount o{ position searching and calculation in coordinates extracting of targets in 2D image, and it can also help identify multiple targets. The results showed that our method could reach an accuracy of 1. 608 mm, which was about one thousandth of the measurable range, and the reconstruction time for 4 targets without geometric constraints is 23.87 ms, namely 41.9 fps.

关 键 词：定位系统 线阵CCD 多目标点 直接线性变换 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]