一种基于点特征的高精度图像配准方法

A High Precision Image Registration Method Based on Point Features

提出了一种基于点特征的多源图像高精度配准方法。通过在遥感影像上提取密集特征点,利用多种方法结合进行严格匹配获得同名点对。通过构建不规则三角网分割影像,再利用仿射变换模型实现每个三角形的配准。实验证明了该方法的有效性和较高的适用性。

This paper put forward a high precision multi-resource image registration method based on point features. It obtained the matching point pairs using multi strict matching method through extracting dense point features on remote sensing images. Then it divided image by constructing triangulated irregular network, and realized the registration of each triangle using Affine Transformation Model. It testified the validity and applicability of the method by experiments.