针对小高比立体匹配当中的亚像素精度和粘合现象问题,提出了一种基于最大似然估计的小基高比立体匹配方法.该方法首先根据混合式窗口选择策略为参考图像中的每一点确定匹配窗口；然后在视差范围内根据规范化互相关函数计算匹配代价,再利用胜者全取策略计算每一点视差；最后采用基于最大似然估计的亚像素匹配方法获得亚像素级视差.实验结果表明,该方法有效地减少了立体匹配中的粘合现象,同时获得了较高精度的亚像素视差,其平均亚像素精度可达1/20个像元.%This paper proposed a small baseline stereo matching method based on maximum likelihood estimation for the sub-pixel accuracy and adhesion phenomenon in small baseline stereo vision. Firstly,it used the hybrid window strategy to determine the matching window for each point in the reference image. Secondly.it employed the normalize cross correlation function to compute matching cost for possible matching points in the disparity range,then utilized the winner-take-all strategy to compute the disparity. Finally, it developed the sub-pixel stereo matching method based on maximum likelihood estimation to attain the sub-pixel disparity. The experimental results show that this method efficiently decreases adhesion phenomenon and achieves more accurate sub-pixel disparity.
展开▼
