Real-time Extraction Of Fire Line Based On Infrared Sequence Images And Fire Line Optimization For Spreading Simulation

Forest fire disturbance is one of the important influences on forest structure and forest environment.Accurate simulation of forest fire spread is an effective way to reduce forest fire hazard.Therefore,it is of great significance to study the occurrence of forest fires under different environmental conditions.In order to simulate the spread process of forest fire accurately,it is necessary to consider the characteristics of fuel,the characteristics of fuel bed and the influence of external conditions.Strong forest fires not only burn down the forest,greatly reduce the forest water and soil conservation,climate regulation function,often cause casualties and property losses.Therefore,it is very important to predict forest fires in time.In this paper,small understory combustibles of typical inflammable tree species in Heilongjiang Province are taken as the research object to reduce the fire hazard.The point burning test was carried out comprehensively,and the whole process of point burning test was observed by low-cost,flexible and convenient UAV with wide field of vision.Firstly,the whole process of the spread video image is preprocessed,and the threshold segmentation algorithm and Canny edge detection algorithm are used to extract the fire line and calculate the spread rate of forest fire.Secondly,Pearson correlation analysis and partial correlation coefficient were used to analyze the correlation of the collected factors affecting the spread of forest fire.Finally,the RothermelHuygens principle model was constructed to simulate the spread of forest fire line,which provided theoretical support and solutions for the in-depth exploration of forest fire spread law and forest fire prevention and control.The main conclusions are as follows:(1)In terms of data collection and actual fire line extraction,compared with the traditional forest fire experiment or field investigation to collect fire spread speed,the use of UAV data collection is more efficient.In addition,Canny operator based on UAV infrared image extraction of fireline contour has an optimal solution for the existing edge extraction.Canny operator extraction of fireline can identify as much as possible the actual edge in the image and be as close to the actual edge as possible.(2)Eight factors affecting forest fire spread were collected in this paper.Through correlation analysis,it was concluded that wind speed,fuel moisture content and slope were the dominant factors affecting the spread rate,which provided the basis for the subsequent construction of forest fire spread model.(3)The Rothermel-Huygens principle model has good applicability in simulating the spread of forest fires.The simulation results of this model are good in indoor and outdoor point burning and can accurately predict and capture the location of fire lines.The accuracy and robustness of the Rothermel-Huygens model were quantitatively analyzed,and it was found that the simulation accuracy of the constructed model reached more than 75% in the indoor point burning test and more than 80% in the field point burning test.This shows that the Rothermel-Huygens model has high accuracy and robustness in simulating the spread of forest firesIn general,this paper proves that Rothermel-Huygens model has good applicability in simulating forest fire spread,and has stable performance in small-scale fire spread,which provides a new scheme and data reference for solving forest fire spread.