A new feature-point-based stereo matching algorithm is proposed in this paper, in order to match feature points more efficiently and accurately.

为了快速并更精确的对特征点进行立体匹配，本文提出了一种新的基于特征点的立体匹配算法。

youdao

This shape recognition method based on real valued boundary points is employed to match two shapes with the proportional property of the arrays of their feature values.

本文提出的图形识别方法基于边缘点的实数值坐标，利用两幅图形特征值数列的比例特性进行匹配。

youdao

Then, normalized cross-correlation matching and probability support matching method are used to match the images feature points.

然后利用归一化互相关匹配法和概率支撑法对特征点进行匹配。

youdao