基于改进Harris-SURF算子的遥感图像配准算法

Remote Sensing image registration based on the Harris-SURF algorithm

针传统SURF算法在遥感图像配准中存在精度和准确率不高的问题,提出基于改进的Harris-SURF图像配准方法。利用高斯二阶导数与原始图像进行卷积,提取具有尺度不变的特征值点,对Harris算法不具备尺度不变性进行改进,利用仿射变化提高配准的精度。最后,利用RANSAC算法剔除配准错误点,完成配准。

To tackle the problems of low accuracy and accuracy in Remote Sensing image registration form traditional SURF algorithm,this paper proposes an improved Harris-SURF algorithm of image registration.The Gaussian second derivative is used to convolve with the original image to extract the scale-invariant feature points. Improving Harris algorithm which does not have scale invariance, the accuracy of registration is improved by affine transformation.At last, the RANSAC algorithm is used to eliminate the registration error points.