Research Of The Technology Of Extracting Forestry Information From Remote Sensing Image

The classification of Remote Sensing image is one of the important research content in RemoteSensing image processing. The classified precision influences the application level and practicalvalue of the Remote Sensing data directly. Recovering forestry information from all kinds of landtypes can be great importance to statistic forestry area in some special region,enforce ecologyconstruction and establish the policy of taking back from agriculture also forest. But at present themost used method in recognizing Remote Sensing image is judging by people. The method does notfit the demand for real time operation, and need more manpower, material resources.According to the project of "Auto recovering technology of Forestry information", this papertakes the SPOT5 Remote Sensing image of zhungeer city in Sep 2005 as source data. Its major tasksare to research the texture analysis and neuron network using in Remote Sensing imageclassification. According to the actual situation of the selected areas, we divide the land cover/useinto six classes: conifer, broadleaf, water, grassplot, everglade, and bottomland. In the experimentbased on texture analysis, we use the energy, contrast, entropy, correlation, inverse differencemoment and DEM also slope as feature values, and adopts neuron network as classifier. We choosefirst band of SPOT5 to get texture feature value, let "winner takes all" as classification rule.According to the method mentioned above, we use the Remote Sensing image classification system developedin vc++ environment to do the experiment. The result shows that method we used can obtain better outcome.