Remote sensing image reconstruction method and device, processor and electronic equipment

Abstract

The invention discloses a remote sensing image reconstruction method and device, a processor and electronic equipment. The method comprises the following steps: acquiring a low-resolution image of a remote sensing image to be reconstructed; the low-resolution image is input into a preset reconstruction network model to obtain a high-resolution image of the to-be-reconstructed remote sensing image, the preset reconstruction network model comprises a cascade residual aggregation module and a boundary perception loss module, the cascade residual aggregation module uses multiple groups of sample images to obtain the high-resolution image of the to-be-reconstructed remote sensing image through weight parameters of a Bayesian convolution kernel trained by deep learning, and the boundary perception loss module obtains the high-resolution image of the to-be-reconstructed remote sensing image; the pixel sensing module is used for sensing a low-resolution image and predicting missing pixels of the low-resolution image through a Bayesian convolution kernel, the boundary sensing loss module is used for determining boundary information of a high-resolution image, and the boundary information is used for modifying weight parameters of the Bayesian convolution kernel. According to the invention, the problem of poor reconstruction result of the remote sensing image in the prior art is solved.

Description

technical field [0001] The present application relates to the field, and in particular, to a method, apparatus, processor and electronic device for reconstructing a remote sensing image. Background technique [0002] The invention relates to the field of satellite remote sensing and deep learning. Aiming at the problem of improving low-resolution remote sensing images, a remote sensing image super-resolution reconstruction method with Bayesian deep learning and boundary perception loss modules is proposed. [0003] Nowadays, in the field of remote sensing, a 0.75m remote sensing image costs 2000-10000 to obtain an image, while a 10m remote sensing image can be obtained free of charge. If there is a super-resolution reconstruction algorithm with high accuracy, 10m can be obtained. The free image reconstruction of its sharpness can save cost to obtain high-definition remote sensing images. However, the current super-resolution reconstruction algorithms mainly have the followi...

AI Technical Summary

Problems solved by technology

[0007] The main purpose of this application is to provide a remote sensing image reconstruction method, device, processor and electronic equipment to solve the problem of poor reconstruction results of remote sensing images in related technologies

Images