Hyperspectral image feature extraction method based on morphology

Abstract

The invention discloses a hyperspectral image feature extraction method based on morphological analysis, and the method comprises the following steps: carrying out the principal component analysis ofa hyperspectral image, and selecting first T principal component components, and T is an integer greater than 0; Constructing a topological tree for each selected principal component; Constructing a topological tree for each selected principal component; Counting the attribute type of each leaf node of each topological tree, and selecting whether to reconstruct the topological tree or not according to the attribute type; Calculating an extinction value corresponding to the attribute value of each topological leaf node, and cutting the topological tree according to the magnitude of the extinction value to obtain a cut topological tree; And reconstructing the shear topology tree as a main component image to obtain extinction profile characteristics. According to the scheme provided by the invention, the extracted image is low in feature dimension and strong in noise interference resistance; And the feature complexity is low and the classification precision is high.

Description

technical field [0001] The invention relates to a remote sensing image feature extraction method, in particular to a hyperspectral image feature extraction method based on morphological analysis, which belongs to the technical field of digital image processing. Background technique [0002] In recent years, hyperspectral imaging has become more and more common for classification, mainly because it contains hundreds of narrow continuous bands, which provide rich spectral information and can provide very accurate classification for some simple scenes. However, for some categories in complex scenes, it is difficult to accurately classify only relying on spectral features, especially for those materials with similar reflectivity, which requires consideration of other information contained in hyperspectral images. With the development of hyperspectral imaging technology and the continuous maturity of airborne and spaceborne imaging methods, the extracted hyperspectral not only ha...

AI Technical Summary

Problems solved by technology

Due to the following deficiencies in the attribute profile: 1) It is difficult to set the optimal threshold; 2) Different images need to set different thresholds; 3) It may get very sparse feature space

Images