High spectral image shadow detection and spectrum recovery method and device

Abstract

The invention provides a high spectral image shadow detection and spectrum recovery method and device. The method comprises steps that S1 based on pre-defined growth criteria, shadow detection on a high spectral image is carried out to acquire a shadow image G; S2, based on the shadow image G, erosion operation and dilation operation are utilized to acquire a non-buffer region image GS and a buffer region image Gbuffer; and S3, a moment matching method is utilized to carry out spectrum recovery of the non-buffer region image GS, and a point spread function method is utilized to carry out spectrum recovery of the buffer region image Gbuffer. The method is advantaged in that disadvantages in the prior art are solved, a shadow detection and recovery problem of the high spectral image is solved, the space and the spectral information of surface features covered by shadow can be more recovered, so image quality is improved, and more effective input for subsequent image processing is provided.

Description

technical field [0001] The present invention relates to the field of image processing, and more specifically, to a method and device for shadow detection and spectral restoration of hyperspectral images. Background technique [0002] At present, the key to distinguish hyperspectral images from traditional RGB images is that the spectral resolution in the imaging band range reaches the nanometer level, which enables the ground objects that cannot be identified in conventional remote sensing to be effectively identified and recognized in hyperspectral remote sensing. Quantitative inversion. [0003] However, in the quantitative inversion of hyperspectral remote sensing, the existence of hyperspectral image shadows seriously weakens the spatial and optical characteristics of the image, which directly affects the accuracy of quantitative inversion. Therefore, how to remove the influence of shadows and restore the spatial and spectral information of hyperspectral images It is on...

AI Technical Summary

Problems solved by technology

[0003] However, in the quantitative inversion of hyperspectral remote sensing, the existence of hyperspectral image shadows seriously weakens the spatial and optical characteristics of the image, which directly affects the accuracy of quantitative inversion. Therefore, how to remove the influence of shadows and restore the spatial and spectral information of hyperspectral images It is one of the main contents of improving quantitative remote sensing inversion

[0004] At present, the detection and compensation of image shadows are often based on RGB images, and these algorithms cannot be directly applied to shadow detection and image restoration of hyperspectral images.

Images