Trinocular Stereo Matching Based on Correlations Between Baselines and Disparities  

作　　者：GUANYepeng GUWeikang 

机构地区：[1]DepartmentofInformationandElectronicsEngineering,ZhejiangUniversity,Hangzhou310027,China

出　　处：《Chinese Journal of Electronics》2004年第4期708-710,共3页电子学报（英文版）

摘　　要：Gray correlation technique is utilized to take the multi-peak feature points with gray correlation coefficients less than a certain range of maximal correlation coefficient as a potential candidate matching set. There exists the maximal correlation between the correct disparities and their corresponding baselines. A trinocular stereo matching algorithm is proposed based on correlations between the baselines and disparities. After computing the correlations between the baselines and disparities, the unique matches can be determined by maximal correlation coefficient. It is proved that the algorithm proposed is valid and credible by 3-D reconstruction on two pairs of actual natural stereo images.

关 键 词：灰色相关 基线 不均匀性 图像处理 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]