武汉工程大学学报2016，Vol.38Issue(4)：386-393,8.DOI:10.3969/j.issn.1674-2869.2016.04.014
双目实时目标三维测量实现方法的研究
Realization Method of Three Dimensional Measurement of Real-Time Target of Binocular Stereo Vision
摘要
Abstract
To improve efficiency of the target-matching, a matching algorithm based on real-time target charac⁃teristics of binocular stereo vision, namely absolute window error minimization(CAEW)was presented. Firstly, the camera calibration was realized by the Zhang's calibration method based on the basic principle of camera, and the final data of binocular stereo calibration was tested by Bouguet algorithm. Then, the target detector was trained by AdaBoost iterative algorithm for better target detection. The evaluated effects of CAEW algorithm was compared with that of the commonly Speeded-up Robust Feature(SURF)algorithm, and the results show that the evaluated effects of CAEW algorithm achieve more than 90%, which is significantly improved, and can meet the requirement of the application of binocular real-time target matching. By comparing with the experimental results of SURF and CAEW algorithm, it demonstrates that the reduction of unnecessary operation can improve the matching speed.关键词
双目立体视觉/特征点匹配/三维测距Key words
binocular stereo vision/feature point matching/3D distance measurement分类
计算机与自动化引用本文复制引用
肖永强,王海晖,刘奥丽,王子维,章刘斌..双目实时目标三维测量实现方法的研究[J].武汉工程大学学报,2016,38(4):386-393,8.基金项目
武汉工程大学大学生校长基金项目（2015047）；湖北高校2015年省级大学生创新创业训练计划项目 （）
