Hyperspectral Remote Sensing Ground Object Classification Method Based on Superpixel Tensor Sparse Coding

Abstract

The invention discloses a hyperspectral remote sensing object classification method based on superpixel tensor sparse coding, which overcomes the shortcomings of the prior art that the spatial information of the hyperspectral image cannot be fully utilized for classification and the classification speed is slow. The steps realized by the present invention are: (1) input hyperspectral image to be classified; (2) construct hierarchical space similarity matrix; (3) obtain superpixel set; (4) construct mark sample dictionary; (5) solve sparse coefficient matrix; (6) superpixel classification; (7) output the classification result of the hyperspectral image to be classified. The invention has the advantages of maintaining the spatial consistency of homogeneous regions of hyperspectral images and fast classification speed, and can be used for rapid classification of hyperspectral images.

Description

technical field [0001] The invention belongs to the technical field of image processing, and further relates to a hyperspectral remote sensing object classification method based on superpixel tensor sparse coding in the technical field of hyperspectral image classification. The invention can be used for fast ground object classification on hyperspectral remote sensing images. Background technique [0002] At present, the focus of research on hyperspectral data is mainly two aspects of dimensionality reduction and classification. Among them, the classification is mainly to process the data according to the different characteristic information of different types of data, and assign a label to each pixel, so as to realize the classification and recognition of the ground objects. According to whether there are labeled samples to participate in the processing process, classification methods can be divided into unsupervised classification methods, supervised classification method...

AI Technical Summary

Problems solved by technology

The disadvantage of this method is that in the process of solving the sparse coefficient, it is necessary to solve the sparse coefficient on the mapped dictionary for each mapped sample. would create a large computational burden

The disadvantage of this method is that it does not take into account the spatial information between adjacent samples, which will cause poor spatial consistency in homogeneous regions and affect the recognition accuracy of hyperspectral images.

Images