Method for detecting fire scene spreading area in forest grassland fire disaster

Abstract

The invention provides a method for detecting a fire scene spreading area in a forest grassland fire disaster, which is used for solving the problem of detecting the spreading area of a fire scene within a certain time length in the prior art.The fire scene spreading area detection method comprises the following steps of firstly, respectively acquiring PMS and IRS files from GF-4PMI images at a T1moment and a T2 moment of a to-be-detected time period, performing radiometric calibration, physical quantity calculation, geometric fine correction, waveband recombination and geometric registration, and cutting to generate 6 waveband image files covering the T1 moment and the T2 moment of a common interested area,and respectively identifying related pixels in the images obtained at two momentsby utilizing an ignition point and a burned area detection algorithm, generating intermediate temporary files Temp1 and Temp2, and detecting a fire scene spreading area.The fire scene spreading area,the average surface speed and the fire head spreading direction in the forest grassland fire disaster occurrence process are quantitatively and dynamically detected, and reliable data are provided forcommanding fire disaster emergency fighting work.

Description

technical field [0001] The invention belongs to the field of natural disaster detection, and in particular relates to a method for detecting the spread area of ​​a forest and grassland fire. Background technique [0002] In the natural state, such as forests and grasslands, due to the existence of fire in the environment and its own flammability, it is very easy to cause wild fires, and the losses caused by this are very huge every year. In the process of emergency fighting after a forest and grassland fire, it is necessary to pay close attention to the spread of the fire scene and effectively dispatch and command the fire fighting work. [0003] Forest and grassland fire spread is an important reference index for forest and grassland fire risk analysis and fire fighting decision-making. The existing forest and grassland fire spread models at home and abroad mainly include the physical mechanism model represented by the Rotherml model in the United States, and the FBP model ...

AI Technical Summary

Problems solved by technology

In the daily satellite remote sensing monitoring of forest and grassland fires at home and abroad, Terra / Aqua MODIS, NOAA / AVHRR, FY-3, etc. are mainly used with medium and low spatial resolution (thermal infrared sensors at the kilometer level) and high temporal resolution (day, night) Due to the limitations of the time, space and radiation resolution of these satellite images, people usually only use the single-time satellite images obtained to extract the fire pixels or fires in the forest and grassland at that moment. The pixels of the ruins, did not detect the spread area of ​​the fire scene within a certain period of time

Images