Multispectral and panchromatic image classification method based on attention mechanism

Abstract

The invention discloses a multispectral and panchromatic image classification method based on an attention mechanism, and belongs to the technical field of image processing. The method comprises the following steps: constructing a training sample set and a test sample set, and carrying out initial feature extraction on multispectral samples and panchromatic samples in the training sample set to obtain initial features of the multispectral samples and initial features of the panchromatic samples with the same feature size; continuously and respectively inputting into the lightweight sharing-separation network for deep feature extraction to respectively obtain deep features of the multispectral sample and deep features of the panchromatic sample; continuing the self-adaptive feature fusion respectively, and outputting fusion features; classifying the fusion features, and training the network according to a classification result; and based on the trained network, obtaining a classification result of multispectral samples and panchromatic samples in the training sample set. The problems of low classification time cost and poor multi-sensor fusion effect in the prior art are solved.

Description

technical field [0001] The invention belongs to the technical field of image processing, and relates to an attention mechanism-based multispectral and panchromatic image classification method. Background technique [0002] The rapid development of aerospace technology has produced a large number of remote sensing images of different sensors, so the effective classification of ground objects combined with multiple sensor images has become a research hotspot, especially for the classification of multispectral and panchromatic images. Generally, a remote sensing satellite has a panchromatic sensor, which can perform spectral response to a wide range of spectra to form a panchromatic image. Panchromatic images are grayscale images with high spatial resolution, but because there is only one spectral band, the spectral resolution is low, and the type of ground objects cannot be determined, which is extremely unfavorable for the identification of ground object types. In order to m...

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the shortcomings of the above-mentioned prior art, the object of the present invention is to provide a multispectral and panchromatic image classification method based on attention mechanism, which solves the problems of classification time cost and poor effect of multi-sensor fusion in the prior art, To balance the classification accuracy and time cost of multispectral and panchromatic images

Images