针对某高分辨率遥感卫星高精度图像配准要求,提出了一种采用边缘特征的提取和特征点提取相结合的方法,解决了卫星云图存在云层遮挡导致图像边缘信息无法准确提取的问题。该方法基于改进的Canny算法和Harris检测算子对图像的边缘特征和特征点进行提取,并采用Hu氏不变矩实现图像特征点的匹配,通过仿真试验验证了该方法对提高图像配准精度的有效性。%In this paper, for the image registration accuracy requirements of a high-resolution remote sensing satellite, an edge feature extraction method combined with a feature point extraction method is proposed. The combined method is used to solve the problem that the edge information of the satel- lite image can not be extracted accurately, as the cover of clouds. The proposed method in this paper is based on the improved Canny algorithm and Harris detection operator for image edge features and feature point extraction, and the use of Hu's invariant moments for matching of image feature points. The effectiveness of the proposed method to improve image registration accuracy is verified by simulation experiments.
展开▼