Remote sensing image space spectrum unsupervised classification method based on multi-objective modality optimization algorithm

Abstract

The invention discloses a remote sensing image space spectrum unsupervised classification method based on a multi-objective modality optimization algorithm. The method comprises: the number of types of remote sensing images is automatically determined through an adaptive differential evolution algorithm; Constructing a spatial information item, and constructing a spatial-spectral combined objective function in combination with Jm; And increasing the optimization of the remote sensing image unsupervised classification result by using a multi-target modality optimization algorithm. The method improves the automation and intelligence of the remote sensing image unsupervised classification algorithm, does not need to manually input the number of priori categories in advance, and comprehensively uses the global and local search capabilities of the modulo optimization algorithm, so that the precision of the unsupervised classification result is improved.

Description

technical field [0001] The present invention is based on the technical processing field of remote sensing images, and in particular relates to a method, system and medium for space-spectrum unsupervised classification of remote sensing images based on a multi-objective meme optimization algorithm. Background technique [0002] Remote sensing classification plays a vital role in remote sensing applications such as land use mapping and precision agriculture. The unsupervised classification of remote sensing images is different from the supervised classification of remote sensing images. It does not require any prior sample information, but only uses information such as the structure of the image itself to perform image classification operations, reducing the selection of a large number of high-precision samples. The application capability of remote sensing images has been expanded. On this basis, the unsupervised classification of remote sensing images has attracted the atten...

AI Technical Summary

Problems solved by technology

[0004] (1) The number of image categories needs to be manually input in advance, rather than automatically determined, lacking in automation;

[0005] (2) The objective function optimization method in the traditional unsupervised classification method belongs to the gradient descent method, which is easy to fall into the local extremum in the high-dimensional and complex remote sensing image data space, that is, lacks the global optimization ability;

[0006] (3) The spatial structure information of the image is not fully considered. Due to the influence of noise and mixed pixels, it is easy to lead to poor unsupervised classification results.

Images