Application of the extraction of the image feature points by improved SIFT algorithm

In the application of the image mosaic, SIFT algorithm mainly focused on feature information of the points. The geometric and statistical information in matched feature points are not concerned, so there are many mismatch feature points in it. Harris corner detection algorithm is a very effective image corner extraction algorithm. In this paper, Harris algorithm is used to obtain more accurate matched feature points. The method can extract more less steady, reliable, uniform and reasonable corner points, and improve the efficiency of image mosaic.