Research On Montoring Method Of Conversion Of Farmland To Forests Of Forest Area Based On SPOT-5 Image

The conversion of farmland to forests for Green Project is implemented in one of six key forestry ecological projects in China, it is to reduce soil erosion, reduce wind and sand disasters, improve the ecological of effective measures.Basic engineering of forest ecosystems can be restored, greatly improved on the middle reaches of the Yangtze River and Yellow River areas of ecological environment. Conversion of farmland to forests has been the effectiveness of project construction, quality is meeting its objectives, the project has to be focused on conversion of farmland to forests issues.With the multi-spectral, high spatial resolution satellite remote sensing data SPOT5 been widely applied to the monitoring of the environment. Explore the actual needs for forestry ecological theory of technology of remote sensing monitoring methods, especially small-class area of remote sensing data to improve estimation accuracy, is yet to be resolved.According to the research, using remote sensing technology to monitor the forests, can save a lot of manpower, and can avoid human factor to impact the test results of returning farmland to forests to the greatest extent.In this study, Combining the characteristics of SPOT5 Data,Liujiang town Hongya County, Sichuan Province as the study area,Use of remote sensing using ERDAS software SPOT5 to pre-processing,image interpretation,image supervised classification,Accessed to forstland area through image SPOT5 reached classification accuracy, the following conclusions:(1) Through the fusion of subjective and objective image evaluation, results of the appraisal:Brovey transform approach the results obtained image quality in its spatial and spectral quality of the best overall performance,suitable for monitoring the use of remote sensing hilly areas.(2) Using supervised classification and unsupervised classification methods to classify the images were processed, and then after the image on the classification accuracy of analysis and evaluation.The results show that:Image classification using supervised classification accuracy was significantly higher than non-supervised classification,able to distinguish between woodland and farmland, the conversion of farmland to forests project of the monitoring process in the use of supervised classification is more accurate.(3) The classification accuracy of the kappa coefficient were 0.9196 and 0.9063 higher classification accuracy.Man-machine interactive interpretation accuracy of 91.20%.Show that this way can be applied to monitor the conversion of farmland to forests.(4) The study area with the design of returning farmland to forest operations andremote sensing images, Liujiang town in 2004 and 2008 the classification accuracy of remote sensing,respectively 91.83 percent and 91.82 percent,it can be used to detect the completion of conversion of farmland to forests.