Spatiotemporal resolution enhancement via compressed sensing

In this paper, we propose a new compressed sensing based approach to enhance the spatial-temporal resolution of the remote sensing images with a pair of time-continuous spatial-temporal images and a low spatial resolution image at the same place. In compressed sensing, the measurement matrix is a key element to success. This paper presents a novel solution space model for designing the measurement matrix by establishing the correspondence between the spatial-temporal image pair to enhance the spatial-temporal resolution. The matrix we get does not only reflect the relationship between the high- and the low-spatial resolution images, but also have high randomness, thus satisfies the reconstruction requirements (e.g., RIP restriction) in compressed sensing. To verify the effectiveness of our method, we give the experimental reconstructed results and compare our results with the traditional Gaussian Random matrix and the Toplitz matrix. The experiment demonstrates the effectiveness and superiority of the proposed method.