Forest fire burned area extraction method and system

Abstract

The invention discloses a forest fire burned area extraction method and system. The extraction method comprises the following steps: acquiring a multispectral remote sensing image of a forest; determining a plurality of spectral index characteristics according to the multispectral remote sensing image; wherein the spectral index characteristics comprise a normalized vegetation index, a shadow vegetation index, a fire passing area identification index, a normalized difference water body index and a global environment monitoring index; acquiring various texture information of the multispectral remote sensing image; acquiring near-infrared band data information of the multispectral remote sensing image; fusing the multiple spectral index features, the multiple texture information and the near-infrared band data information by adopting a band fusion mode to obtain a fused multi-band image; and carrying out unsupervised classification on the fused multi-band images by adopting a binary K-means clustering algorithm, and extracting patch information of the burned area. According to the method, the influence of a single spectral index and spectral similar ground objects on the burned areaextraction result can be avoided, and the burned area recognition accuracy is improved.

Description

technical field [0001] The invention relates to the field of forest fire monitoring, in particular to a method and system for extracting forest fire burnt areas. Background technique [0002] The remote sensing monitoring technology of forest fire burnt areas can quickly estimate the fire area information, assess the social and economic losses, and avoid a large number of field investigations. According to the different satellite remote sensing data obtained, the forest fire burnt area monitoring technology based on satellite remote sensing can be divided into two types. The first is to use pre-disaster and post-disaster multi-temporal visible light / near-infrared remote sensing data to identify and extract burnt areas, based on the changes in the spectral characteristics of ground vegetation coverage before and after the fire, mainly using remote sensing image change detection methods. The second is to use post-disaster single-phase visible light / near-infrared remote sensin...

AI Technical Summary

Problems solved by technology

However, there are the following problems in using post-disaster single-temporal satellite multispectral remote sensing images to monitor burnt areas: (1) A single spectral index is easily affected by non-fire causes such as vegetation seasonal changes, drought and water shortage, which can easily cause burnt areas (2) The phenomenon of "same spectrum and different objects, same object and different spectrum" brought by high-resolution satellite imagery will lead to wrong and missed extraction of pixels in the burnt area, especially cloud shadows or terrain shadows, and darker colors. Deep water bodies and bare soil of certain colors are often very similar to burnt patches, so credible burnt information cannot be obtained

Images