The Application Of High-resolution Remote Sensing Data In The Project Of Returning Farmland To Forests

The project of Returning Farmland to Forests is an important ecological project. The project change soil erosion serious farmland, desertification of farmland and slope sloping farmland to forests, in order to improve the ecological environment and promote the rural industrial structure adjustment. Dynamic monitoring, we can get the project progress and better know the progress of the works on the Returning Farmland to Forests Project.In this study, high-resolution remote sensing images were used to monitor the project of returning farmland to forests in Zhangjiajie National Forest Park. Remote sensing images used in this paper are the SPOT-5images which were acquired in2008and the GeoEye-1images which were acquired in2009. We could achieve available images which are used for information extraction by characteristics of the remote sensing data, band combination, image fusion, orthorectification and cutting processing. We used software developed based on GE to extract high-precision elevation data on Google Earth, and DEM can be produced in the geographic information system. Conducted a comparative analysis of the five kinds of fusion methods of GeoEye-1remote sensing images, we draw the following conclusions: IHS transform has a good effect on the increase in the amount of information, on the original multi-spectral data maintained and reflecting the texture of space, and is suitable for the monitoring. Four remote sensing image classification methods are used for information extraction. The face object-based classification is better than the pixel-based classification method, and the classification accuracy can be increased to more than90%.First, both high-resolution remote sensing image are auto-classification by computer with the method of the face object-based classification and then to extract the forest resource change information. The result is pre-classification accuracy of96.53%, post-classification accuracy of97.12%, the pre-forest coverage of96.4%, post97.35%. Contrast to the two remote sensing images of forest land area changes:2008to2009, Zhangjiajie National Forest Park new woodland area of46hectares and Zhangjiajie National Forest Park on the17th compartment, ranking the18 compartment first. Using DEM overlay2009remote sensing images, we can find the Zhangjiajie National Forest Park sloping farmland is24.29hectares. Distribution of arable land on the slope analysis, we can see that part of the compartment is no longer slope farmland.Slope farmland mainly concentrated in the lth compartment.In conclusion the park returning farmland to forests has been basically completed, the forest coverage has reached more than97%, and the next phase of work is to protect the achievements already made.