Hyperspectral coastal wetland sub-pixel change detection method based on cooperation of clustering segmentation and coupling end member extraction

Abstract

The invention relates to a hyperspectral coastal wetland sub-pixel change detection method based on cooperation of clustering segmentation and coupling end member extraction. The method comprises thefollowing steps: step 1, preprocessing two-time-phase hyperspectral data; and step 2, performing multi-feature extraction on the spliced images by utilizing the spectral features, the geometrical features and the texture features. The method has the beneficial effects that firstly, clustering segmentation is carried out on image categories through multi-feature extraction to obtain a ground objectsegmentation map, and ground object types are finely segmented through spectral features, geometrical features and texture features of ground objects by utilizing hyperspectral remote sensing data; fine classification research is carried out on different end members of ground objects of the same kind: fine end members with high purity are extracted by using a coupled non-negative matrix factorization method, then end member abundance is solved based on a sparse unmixing method, and coastal wetland change conditions are detected through abundance of two time phases. And fine change detection of the coastal wetland on the sub-pixel scale is realized. And the method has simple operability and high practicability.

Description

technical field [0001] The invention belongs to the technical field of remote sensing image applications, and in particular relates to a hyperspectral coastal wetland sub-pixel change detection method in which clustering segmentation and coupling endmember extraction are coordinated. Background technique [0002] In recent years, with the rapid development of remote sensing technology, hyperspectral remote sensing has been widely used due to its rich and fine spectral information. Different from change detection based on multispectral data, the potential change information of ground objects can be further explored through hyperspectral images. However, due to the limitations of hyperspectral sensor hardware, while obtaining hyperspectral resolution, its spatial resolution is generally low. Therefore, during the imaging process of hyperspectral sensors, two or more ground objects appear in the same pixel at the same time. Phenomenon, which is a mixed pixel, greatly affects th...

AI Technical Summary

Problems solved by technology

At present, due to the diversity of coastal vegetation, there are many problems in the decomposition of its mixed pixels, including the following points: First, due to the low spatial resolution of the hyperspectral image itself, there are often a large number of mixed pixels in the obtained coastal wetland images ; Second, the comple

Images