Image semi-supervised classification method and device and computer readable storage medium

Abstract

The invention relates to the technical field of high-dimensional image processing, in particular to an image semi-supervised classification method and device and a computer readable storage medium. The invention provides an image semi-supervised classification method. The method comprises the following steps: reading hyperspectral image cube data and a training set; rearranging the hyperspectral image cube data into a data matrix; calculating a full connection weight matrix; calculating k neighbors; forcing a sparse connection matrix of the nearest neighbor attribute; constructing a sparse connection matrix of mutual k neighbor attributes; obtaining a sparse connection matrix with complementary neighbor attributes; sparsifying the weight matrix by using a sparse connection matrix with complementary neighbor attributes to obtain a complementary neighbor sparse weight matrix; constructing an initial label matrix; and implementing semi-supervised image classification to obtain a classification result of all pixels of the hyperspectral image cube data.

Description

technical field [0001] The present application relates to the technical field of high-dimensional image processing, in particular, to an image semi-supervised classification method, device and computer-readable storage medium. Background technique [0002] A hyperspectral image is an image cube with high spectral dimensions acquired by a hyperspectral imager. Its spatial plane contains a large number of pixels representing ground object pixels, and each pixel is composed of an approximately continuous spectral feature. The spectral feature is essentially the reflectance of the feature object to light of different wavelengths, and reflects the specific material and properties of the feature, so it can be used to determine the category of the feature corresponding to the pixel. The large amount of data and high-dimensional features of hyperspectral images have brought great challenges to image processing and classification tasks. In addition, since the spectral characteristic...

AI Technical Summary

Problems solved by technology

The hyperspectral image semi-supervised classification method based on the forced k-nearest neighbor graph is very sensitive to the graph structure. The similarity between samples and the graph sparse method will greatly change the classification results. The classification results have single neighbor attributes and strong connection noise. , prone to defects such as isolated points, resulting in low accuracy of image classification results

Images