遥感多图像配准中自动提取特征点的并行算法

A Parallel Algorithm of Automatically Selecting the Characteristic Points in the Multiple Image Registration for Remote Sensing

在遥感多图像配准过程中特征点的选择是一个关键的步骤。实际处理的遥感图像一般很 大,选择足够的特征点又要保证配准的精确度是极其耗费时间的。本文提出了一种用于解决特征点选 择的高性能算法,并将该算法进行并行优化,既保证了特征点的数量,又减少了所需时间。

Selecting the characteristic points is one of the critical steps during the image registration for remote sensing. When the image size is very large, selecting enough characteristic points is time-consuming. In this paper, we introduce a high-performance algorithm that is applied in selecting the characteristic points, and optimize the parallel algorithm. So enough characteristic points can be selected in a short time using this algorithm.