An automatic extraction method of point source risk sources using a time-spectrum-space integrated feature model

Abstract

The invention discloses an automatic point source risk source extraction method using a time-spectrum-space integrated feature model, comprising: step 1, generating a key spectrum-space sub-cube; step 2, extracting a spectrum-space target sub-cube; step 3, establishing a time spectrum Space-integrated feature model to extract sub-point source risk sources; step 4, sub-point source risk source areas are merged to realize automatic extraction of point source risk sources; the present invention effectively and fully utilizes time, spectrum, and space data to establish integrated features model to more accurately and effectively realize the automatic extraction of the shape and type of point source risk sources; establish a time-spectrum-space integrated feature model of hyperspectral remote sensing images, which fully considers the temporal, spatial and spectral characteristics of point source risk source remote sensing data The overall relationship between them, analyzing the probability graph model relationship of their interaction and mutual influence, effectively realizes the automatic extraction of point source risk sources, and improves the accuracy and automation of point source risk source automatic extraction.

Description

technical field [0001] The invention relates to the field of target extraction of hyperspectral remote sensing images, in particular to a point source risk source automatic extraction method using a time-spectrum-space integrated feature model. Background technique [0002] The inspection of the ecological environment of water sources needs to monitor all pollution sources or risk sources that have a direct or indirect impact on the water body of the water source. Among them, point source pollution is a pollution source that has a serious impact on the water body of the water source. The key monitoring object of pollution. Use hyperspectral remote sensing data to automatically extract point source risk source information in water source protection areas, and realize spatial positioning and accurate statistics of all point source risk sources in water source areas, providing favorable technical means for on-site monitoring of water source point source risk sources. [0003] ...

AI Technical Summary

Problems solved by technology

[0004] The invention provides an automatic point source risk source extraction method using a time-spectrum-space integrated feature model, which adopts a time-spectrum-space integrated feature model, fully utilizes the time-phase, space, and spectrum information of remote sensing images, and improves The accuracy and automation of hyperspectral remote sensing image point source risk source extraction are used to solve the existing hyperspectral remote sensing image point source risk source extraction technology that still requires manual participation, and the time, spectrum, and space data are studied separately. Making full use of hyperspectral remote sensing image map integration problem

Images