Geostationary orbit satellite image forest fire detection method and system and storage medium thereof

Abstract

The invention relates to a mountain fire detection method, in particular to a geostationary orbit satellite image forest fire detection method and system and a storage medium thereof. The method comprises the steps of recognizing and extracting suspected fire points in an infrared band through a gray level image analysis method on the basis of the incidence relation of infrared band brightness temperature and gray level number; and secondly, performing fire point identification feature coding on the infrared and multispectral waveband images in a spectral superposition mode on multispectral image information in the same geographic area obtained through geographic coordinate mapping in a non-local depth convolution feature coding mode, and finally realizing discrimination and analysis of real fire points. Waveband physical inversion is used and a fire point rough judgment is made by fully referring to the fire point brightness temperature characteristic of the infrared waveband, and precise fire point spatial positioning of the remarkable forest fire smoke characteristic under the high spatial resolution is completed by utilizing the remarkable texture information of the high-resolution multispectral waveband.

Description

technical field [0001] The invention relates to a mountain fire detection method, in particular to a geostationary orbit satellite image mountain fire detection method, system and storage medium thereof. [0002] Background of the invention [0003] my country's Yunnan region has complex landforms, many high mountains, dense forests, and uneven terrain, which can easily cause forest fires and pose a major threat to the life and property of the country and the people. Traditional mountain fire monitoring methods are mainly carried out manually, with poor monitoring efficiency and limited coverage. narrow. The remote sensing images of geostationary satellites represented by Gaofen 4 have many advantages such as high spatial resolution, sensitive infrared brightness temperature characteristics, wide observation range, and high positioning accuracy. Insufficient important technical auxiliary means are of great significance to the effective monitoring of fire hazards in areas with...

AI Technical Summary

Problems solved by technology

[0006] In order to solve the shortage of imaging spectral resolution of the high-resolution, multi-spectral band geostationary satellite images represented by Gaofen-4 satellite in the current mountain fire detection method, and the number of spectral bands for fire point brightness temperature analysis less, direct spectral calculation and analysis is difficult, and the accuracy of spatial range and position estimation is not high. The purpose of the present invention is to improve the fire point spatial positioning accuracy and recognition accuracy of the fire point detection algorithm. For this reason, the present invention provides a geostationary Method, system and storage medium for mountain fire detection in orbiting satellite images

Images