一种基于稀疏表达的遥感影像时空融合方法

A Spatiotemporal Fusion Algorithm based on Sparse Representation for Remote Sensing Imagery

由于受传感器硬件的限制,遥感影像存在空间分辨率和时间分辨率不可兼得的现象,因此,时空融合技术很重要。针对时空融合中不同传感器之间分辨率存在的较大差异问题,文中采用基于稀疏表达的超分辨率重建方法,构建中间分辨率影像,建立耦合字典模型,采用差分稀疏表达的方法进行时空融合。实验结果表明,该方法比传统稀疏表达方法在ERGAS指标上提高将近10%。

There exists a tradeoff between the spatial resolution and temporal resolution of the remote sensing images because of the limitation of the hardware,Therefore,the spatiotemporal fusion is important in image anlysis. To overcome the large difference of spatial resolution between different images,this paper proposed a novel spatiotemporal fusion algorithm based on superresolution reconstruction,used the single image superresolution reconstruction technique to construct the intermediate resolution images,and then these intermediate resolution images are applied to construct the coupled dictionary model. The high spatial resolution and high temporal resolution images are predicted based on these coupled dictionary model and sparse representation theory. Experimental results demonstrate that the proposed approach can raise the ERGAS index by 10%.