论文部分内容阅读
针对大幅面遥感影像的配准需求,提出了一种带有几何约束的大幅面遥感影像快速自动配准方法。该方法首先结合遥感影像空间地理信息构建影像粗配准模型;然后在粗配准模型的约束下,通过对大幅面遥感影像进行规则格网划分,结合改进的SIFT算法和并行策略实现影像精配准。实验结果表明,本文方法能提高影像的配准效率,且配准精度较高,适用于大幅面遥感影像之间的配准。
Aiming at the registration requirements of large format remote sensing images, a fast automatic registration method for large format remote sensing images with geometric constraints is proposed. In this method, the rough image registration model is constructed based on the spatial geographic information of remote sensing image. Then, the coarse grid image is classified by the regular grid partition of the large-scale remote sensing image under the constraint of the coarse registration model, combined with the improved SIFT algorithm and parallel strategy quasi. The experimental results show that the proposed method can improve the registration efficiency of images, and the registration accuracy is high. It is suitable for registration between large-format remote sensing images.