一种基于改进SIFT算法的图像配准方法

Image Match Method Based on Improved SIFT Algorithm

摘要: 给出一种改进的SIFT算法，并将该算法应用到图像配准中.首先，该改进算法结合Canny算子，去除影响匹配的边缘点.然后采用最近邻与次近邻之比来对96维描述子进行匹配.最后，采用随机抽样一致性（RANSAC）方法消除误匹配.实验结果表明，与原SIFT算法相比，该算法能够有效提高图像配准确度，并减少了10%左右的运算时间.

Abstract: This paper presented a improved SIFT algorithm,and applied this algorithm in image matching.Firstly,integrated the Canny algorithm into the improved algorithm,excluding the edge points.Then,ratio between the first and the second closest distance is used to match the 96-dimensional vectors.Finally,eliminating the inaccurate matching by using RANSAC method.The experimental results show that,compared with the original SIFT algorithm,the modified algorithm can effectively improve the precision of image matching,and the matching time is decreased about 15%.