二值图中测距方法的研究与实现  

Study and Realization of Measuring Distance in Binary Images

作　　者：匡金骏[1] 张杰[1] 

机构地区：[1]成都信息工程学院通信工程系,四川成都610225

出　　处：《成都信息工程学院学报》2007年第z1期62-66,共5页Journal of Chengdu University of Information Technology

摘　　要：详述了Blob分析、模板匹配、背景估计、摄像机标定处理等二值图测距的关键技术。提出了具有代表性的4种二值图测距系统的框架结构。在大量仿真实验结果基础上,分析了影响每种视觉定位系统的性能因素。针对使用模板匹配的系统,提出了改进的区域匹配算法,其复杂度低、鲁棒性高,具有实用价值。The core techniques of Computer Vision which includes Blob analysis,2D correlation,Background Estimate and Camera Calibration are described.This paper classified the solutions into movement and shape.Not only the principles and flow charts of designing four kinds of vision orientation systems are presented but also the factors affecting the performance of the systems based on the simulation results are pointed out.

关 键 词：BLOB分析 区域匹配 运动特征 形状特征 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]