A hyperspectral image emissivity spectrum extraction method based on endmember abundance inversion

Abstract

The invention provides a hyperspectral image emissivity spectrum extraction method based on endmember abundance inversion. First, the whole infrared hyperspectral image is decomposed into mixed pixels to obtain the pure endmembers in the image and the corresponding endmembers. The abundance value of the target, and then obtain the radiance spectrum curve of each pixel point in the image, and then perform curve integration on the radiance spectrum of the target of interest within the set temperature threshold range, and estimate the infrared hyperspectral emissivity according to the integrated area The absolute black body temperature simulated by the curve, and then the emissivity spectrum curve is obtained. When extracting the radiance curve of the target of interest, the present invention considers the influence of mixed pixels, obtains the real radiance curve of the target through end-member extraction and abundance inversion, improves the extraction accuracy of the radiance curve, and then makes the emissivity The extraction accuracy of curves has been improved.

Description

technical field [0001] The invention belongs to the technical field of infrared hyperspectral target characteristics and target recognition, and in particular relates to a hyperspectral image emissivity spectrum extraction method based on endmember abundance inversion. Background technique [0002] Infrared hyperspectral remote sensing is a new technology developed in recent decades. In infrared hyperspectral remote sensing, the primary task is to retrieve temperature and emissivity. Emissivity is a physical quantity that characterizes the ability of an object to radiate to the surrounding environment. The emissivity spectrum of an object is related to its composition and structure. Therefore, in many applications, objects can be identified by the shape of the emissivity spectrum curve. The extraction of emissivity spectral features requires temperature and emissivity separation (TES) of infrared hyperspectral images. [0003] The temperature and emissivity separation meth...

AI Technical Summary

Problems solved by technology

[0006] The object of the present invention is to: aim at the above-mentioned defects in the prior art, propose a hyperspectral image emissivity spectral extraction method based on endmember abundance inversion, and solve the For the problem of mixed pixels, the real radiance curve of the target is obtained through end-member extraction and abundance inversion, which improves the extraction accuracy of the radiance curve, thereby improving the extraction accuracy of the emissivity curve

Images