首页 | 本学科首页   官方微博 | 高级检索  
     

基于双目视觉系统的三维目标识别研究
引用本文:张驰,靳伍银,刘昊,冯瑞成.基于双目视觉系统的三维目标识别研究[J].机床与液压,2010,38(1).
作者姓名:张驰  靳伍银  刘昊  冯瑞成
作者单位:兰州理工大学机电工程学院,甘肃兰州,730050
基金项目:国家自然科学基金(30670529)资助;;甘肃省自然科学基金项目(3ZS042-B25-019)
摘    要:双目视觉系统一般是由一个或两个摄相机从不同角度获取周围景物两幅数字图像,基于人类视觉的视差原理重建周围景物的三维几何信息。通过开源计算机视觉库(OpenCV)重建周围景物的三维几何信息,并提取三维目标的特征信息。采用支持向量机(Support Vector Machine)的软件包LIBSVM进行小样本的学习,从而达到对特定目标的识别。

关 键 词:三维重构  模式识别  视差  OpenCV  支持向量机  

Identification Technique for 3D Target Based on Binocular Visual System
ZHANG Chi,JIN Wuyin,LIU Hao,FENG Ruicheng.Identification Technique for 3D Target Based on Binocular Visual System[J].Machine Tool & Hydraulics,2010,38(1).
Authors:ZHANG Chi  JIN Wuyin  LIU Hao  FENG Ruicheng
Affiliation:School of Mechanical and Electronic Engineering;Lanzhou University of Technology;Lanzhou Gnasu 730050;China
Abstract:The theory of binocular visual system is that two digital images of surrounding scenery are got by one or two cameras from different point of view,then 3D information of surrounding scenery is reconstructed based on the principle of binocular visual system.One identification technique for 3D target was proposed.The characteristic information of target was extracted and reconstructed from two digital images based on Open Source Computer Vision Library(OpenCV).The recognition of given target was realized base...
Keywords:3D reconstruction  Pattern recognition  Disparity  OpenCV  Support vector machine  
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号