Nuclear coordinated expression-based hyperspectral image classification method

Abstract

The invention discloses a nuclear coordinated expression-based hyperspectral image classification method. The method comprises the following steps: carrying out feature selection by adopting a waveband selection strategy with strong operability; carrying out local binary-pattern space feature extraction on the basis of selected different feature groups; carrying out nuclear coordinated expression classification; and finally fusing classification results corresponding to the groups of features and a residual-level fusion strategy and obtaining the final high-precision classification result. According to the method, the textural features of data extracted by a LBP operator are combined, the LBP has the remarkable advantages of rotation invariance and grey level invariance, and the calculation is simple, so that the robustness of the features to be classified is further increased. Finally, a nuclear coordinated expression classifier is used for classification, the calculation efficiency is better than that of the traditional sparse manner and the nonlinear space data can be classified, so that the application range is wider and the application performance is more excellent.

Description

technical field [0001] The invention belongs to the technical field of remote sensing image processing, and mainly relates to land cover classification technology, in particular to a hyperspectral image classification method based on kernel cooperative expression. Background technique [0002] Hyperspectral remote sensing is a kind of high-dimensional information acquisition technology. While acquiring images of surface space objects, it can also obtain continuous and very narrow spectral information of each surface object, that is, to integrate image dimension and spectral dimension information into one. It is rich in Spatial, radiometric and spectral information of surfaces. Compared with broadband remote sensing, hyperspectral data can effectively capture the spectral characteristics of ground objects, and greatly improve the ability to express and identify fine information of ground objects. Object classification is an important application method of hyperspectral remot...

AI Technical Summary

Problems solved by technology

[0004] From a technical point of view, the existing research on hyperspectral image classification methods is mainly focused on spectral space and feature space, and the research on the correlation of image space is not deep enough.

Moreover, most of the studies that focus on feature extraction and feature space classification of remote sensing images have not done in-depth research and discussion on issues such as feature selection that can improve classification accuracy.

Images