一种改进的特征点匹配算法

特征点匹配是计算机视觉中的关键步骤，在很多领域中都有着的重要应用。通过对当前图像特征点匹配方法的研究，提取一种基于特征点的灰度量和几何特征量相结合的匹配方法。该方法首先用Harris算法提取特征点；然后用极线约束减少搜索范围；最后用特征点的灰度量实现特征点匹配。该方法利用极线约束，克服了用灰度量进行特征点匹配计算量大的缺点。提高了匹配速度。实验表明，是一种准确快速的特征点匹配方法。

An Improved Feature Points Matching Algorithm

Feature points matching is the Through research current methods of feature key step in computer vision,which have important applications . points maching, a maching method, which based on the gray value and the geometric characteristics of the point, is proposed in the paper.First,it using Harris operator ex- tracts feature point.Second,it reduces the search area by using epipolar constraint.Last,it uses the gray value of the feature point to achive maching. This method overcome the shortcoming of a large quantity calculation. The experiment results show the approach is accurate and fast.