三维机器视觉中图象匹配的几种摄象机几何模型和误差分析(英文)  

SEVERAL CAMERA GEOMETRY MODELS AND ERROR ANALYSIS FOR IMAGE MATCHING IN 3-D MACHINE VISION

作　　者：刘柏高[1] 袁立行[2] 郑南宁[2] 舒服[3] 

机构地区：[1]西北工业大学电子工程系,西安710072 [2]西安交通大学人工智能研究所,西安710049 [3]湖南省湘雅医院设备科,长沙410000

出　　处：《光子学报》1997年第8期737-741,共5页Acta Photonica Sinica

摘　　要：三维视觉中的一个主要问题是从不同的角度来确定物体的几何位置，而物体的位置则可由两幅实体图象的视差来确定．本文重点讨论图象匹配中几种摄象机的几何模型和误差分析．An important task in stereo machine vision is to determine the three-dimensional location ofobjects from different views(images)of the object scene. The actual location of a scene element can bedetermined from the disparity of two (each in deferent image ) of its depicted entities. Prior toestablishing disparity,however,the correspondence problem (or matching problem) must be solved. Inthis paper,we concentrate on the problem of camera geometry modeling. We discuss several kinds ofcamera geometry for image matching in 3D machine vision. And their error analysis is also studied.

关 键 词：立体模型 三维 计算机视觉 图象匹配 摄象机 

分 类 号：TN948.41[电子电信—信号与信息处理] TP391.41[电子电信—信息与通信工程]