Remote sensing image panchromatic sharpening method based on generative adversarial networks

Abstract

The invention relates to a remote sensing image panchromatic sharpening method based on generative adversarial networks (GAN). The remote sensing image panchromatic sharpening method based on generative adversarial networks includes the steps: 1) constructing a data set: performing target classification and marking on remote sensing image data, and constructing an original image data set includingpanchromatic images and spectral images; 2) designing to a GAN model and a discrimination network model, wherein the GAN model G of GAN learns the random noise vector and the images in the data set and trains a mapping to the generated sample image y, and the sample image generated from the trained G cannot be discriminated as false by the discrimination network model while the discrimination network model D uses the convolutional neural network, and after training, the discrimination network model D can discriminate true and false of the generated data, thus being able to preferably completediscrimination of the classification problem of true or false of generated sample images; and 3) training the GAN model.

Description

technical field [0001] The invention relates to the fields of remote sensing image processing, image fusion, deep learning, etc., in particular to a fusion method of panchromatic images and multispectral images in the field of remote sensing, and belongs to the technical field of remote sensing image fusion. Background technique [0002] Remote sensing images with high spatial resolution and spectral resolution have obvious practical significance for object detection, geographic mapping and environmental monitoring. However, due to the limitations of signal transmission bands and imaging sensor storage, most remote sensing satellites only provide multispectral (MSI) images with low spatial / high spectral resolution and panchromatic (PAN) images with high spatial / low spectral resolution. The two images are highly complementary. Taking advantage of the differences and complementarities of the two images, they are fused into a fused image with clear spatial details and rich spe...

AI Technical Summary

Problems solved by technology

However, due to the limitations of signal transmission bands and imaging sensor storage, most remote sensing satellites only provide multispectral (MSI) imag

Images