基于模板核灰度的Susan特征向量匹配算法

Matching Algorithm Based on Susan Feature Vector of Gray Template

Susan算子是一种基于灰度的特征点获取方法,适用于图像中边缘的检测、角点的检测,论文研究了图像旋转,图像平移和阈值大小对Susan特征点匹配的影响。结果表明,图像旋转和图像平移对特征点的影响不大。当阈值变小时,匹配特征点的效果变佳。该方法处理图像容易实现,对噪声的处理能力强。而且检测目标的速度较快,比较适合应用于图像特征提取中。

Susan operator is a method of obtaining feature points based on gray, and widely used to detect the image edges and corners. The effect of image rotation, image translation and threshold size on matching of Susan feature point is studied in the paper. The results show that image rotation and translation have little effect on the matching. When the threshold becomes smaller, the matching effect becomes better. This method is simple to implement,fast processing speed, high noise immunity and so on, especially suitable for image feature extraction.