针对SAR图像特征点匹配问题,提出了一种基于Facet模型的特征点检测方法。首先基于Facet模型对局部区域做灰度曲面最佳拟合,然后计算拟合曲面中心点的二阶方向导数,取二阶方向导数极大值小于零的点作为潜在特征点,最后通过对极大值的绝对值归一化和局部非极大值抑制提取特征点。实验结果表明,该算法可有效检测特征点,算法的实时性优于传统的SIFT算法。%For the feature point matching of SAR images used in navigation system ,a feature point extraction approach is proposed based on Facet model .Firstly,the image intensity surface is well fitted through the Facet model .Then the second-order directional derivative of the center point of the fitting surface is calculated out,and the point whose second-order directional derivative is less than zero is taken as the potential feature point .Through normalization of absolute values for maximum values of the potential feature points and local non-maxima suppression ,the feature points are obtained .Experimental results show that the proposed algorithm can successfully detect the feature point ,and the real-time performance of the algorithm is better than that of the traditional SIFT algorithm .
展开▼