Remote-sensing image multi-scale object-oriented classification method based on joint sparsity representation

Abstract

The invention discloses a remote-sensing image multi-scale object-oriented classification method based on joint sparsity representation. The method includes the steps of firstly, mining a spatial characteristic of a remote-sensing image and constructing an augmentation characteristic through the combination of the spatial characteristic and a spectral characteristic, secondly, constructing a redundant dictionary by means of a training pixel sample and the augmentation characteristic, carrying out joint sparsity representation on initially segmented patches by the adoption of the redundant dictionary, thirdly, carrying out homogeneity analysis and reconstruction effect analysis on the patches on the basis of the joint sparsity representation, and finally, judging whether patch segmentation is reasonable according to analysis results of the homogeneity analysis and the reconstruction effect analysis, and carrying out classification identification on the patches meeting the requirements of the homogeneity level and the reconstruction effect. The method achieves organic combination of the segmentation process and the classification process so as to obtain proper ground object patches for classification, achieves classification identification of the remote-sensing image on the object level, can obtain a classification result meeting the requirement of visual interpretation, greatly improves interpretation precision of the remote-sensing image, and has great application value.

Description

technical field [0001] The invention relates to the technical field of optical remote sensing data analysis, in particular to a multi-scale object-oriented classification method for remote sensing images based on joint sparse representation. Background technique [0002] With the rapid development of remote sensing technology, especially the launch of high spatial resolution remote sensing satellites in recent years, satellite remote sensing has been more and more applied to various fields of science and production, such as digital city construction, large-scale resource and environmental investigation, environmental monitoring , precision agriculture, archaeology and other special remote sensing monitoring. High-resolution remote sensing data itself has the following characteristics: 1) The amount of single image data has increased significantly; 2) The number of imaging spectral bands has decreased; 3) The geometric structure and texture information of ground objects are m...

AI Technical Summary

Problems solved by technology

[0006] This type of technology excavates the spatial information of high-resolution remote sensing images and improves the interpretation accuracy to a certain extent, but there are still the following problems: 1) The relationship between the size of the pixel and the category of objects and the area of ​​​​the plot is not considered; 2) The image internal The spectral variability produces "salt and pepper" misclassification; 3) Pixel-by-pixel processing takes time

Due to the high spatial resolution, large noise, and large scale span of high-resolution remote sensing images, the selection of a single optimal patch scale has considerable limitations, which is likely to cause over-segmentation or under-segmentation, thereby affecting the image object. Interpretation

On the other hand, manually adjusting the segmentation scale requires users to have certain professional knowledge, which intensifies the difficulty of automatic processing of object-oriented classification technology.

Images