Design And Implementation For Satellite Remote Sensing Forest Fire-point Automatic Monitoring System

Forest fire-point monitoring is a routine operation of weather service. Generally it's a hard and inefficient work because of its basing on a manual and computer aided processing method. In this paper, the study is focused on the automatic remote sensing geographical registration and fire-point detecting method to solve the problem of low level of automation in the fire monitoring operations. Finally Satellite Remote Sensing Forest Fire-point Automatic Monitoring System is designed.As the remote sensing data has a relatively fixed resolution, small geometric distortion, and stable quality, a geographical registration method based on gray correlation and ground control point database is studied to replace inefficient manual registration and the result of an automatic registration experiment performed on 16 examples of EOS/MODIS data which includes day and night in February-may 2008 using the method, indicates that the average registration error on day data is about 0.839 pixels and 0.319 pixels on night data. Basing on a perfect ground control point database, gray correlation is simple, effective and economic.On this basis, we used a simple common fire monitoring and fire area estimation method so that it can be used on both the day and night remote sensing data. Finally, The Satellite Remote Sensing Forest Fire-point Automatic Monitoring System is established by computer programming. The system uses structured and object-oriented programming techniques, improved scalability and stability. In addition, the key algorithm for geometric correction is optimized to improve the processing speed. The system has functions about remote sensing image display, automatic geographical registration, automatic fire-point monitoring, automatic generating of fire-point SMS, and so on.The results of the fact-finding shows the fire monitoring results has a high accuracy.