Super-resolution Reconstruction Method for Single Remote Sensing Image Based on Residual Network and Texture Transfer Model
Remote sensing images are highly susceptible to environmental and weather factors,resulting in reduced resolution.In order to improve the quality of extracting the detailed information from a single remote sensing image,a super-resolution reconstruc-tion method for a single remote sensing image is optimized and designed with the support of fusion residual networks and texture trans-fer models.By considering the degradation phenomenon of remote sensing images,the single low resolution remote sensing image is obtained according to the principle of remote sensing imaging.The assignment of initial remote sensing images is achieved through the steps of defogging,smoothing filtering,and color space conversion.By integrating the residual networks and texture transfer models,the single remote sensing image texture feature labeling is performed to determine the image reconstruction pattern.After the image detailed loss is compensated,the super-resolution reconstruction results of the single remote sensing image are obtained.The remote sensing images of different sizes are taken as a research objective,experimental results show that the proposed method improves the peak signal-to-noise ratio of the reconstructed image by about 204,and the resolution of the reconstructed image is always 1 080 dpi.Also,it significantly reduces the time cost of the image reconstruction.
residual networktexture transfer modelremote sensing imagesimage reconstructionsuper resolution
万方数据
维普
