Monitoring Of Urban Ecological Land Research Based On GF-1 Image

With the deepening of urbanization, internal structure adjustment of urban land change, urban scale has been expanded extension. A large number of ecological land occupation of land for construction, environment and quality of urban ecology continued to decline. Currently, improve land development program, urban ecological environment protection became new urban development goals. Ecological land use remote sensing technology for real-time monitoring of ecological land protection, construction of the new urbanization is important. 2013 success of "GF-1" satellite launch can provide important data to support the process of urban land in urban ecology. But the research is still insufficient, the paper based on the perspectives on the applicability of high variable ratio image of urban ecological monitoring, it will promote the use of high spatial image data to improve monitoring accuracy urban ecological land, which has a very important theoretical and practical value.On the basis of a clear connotation of urban ecological land, select Shijiazhuang City as the research object. GF-1 remote sensing image as the main data source in 2013, it focuses on the applicability of the traditional classification methods and object-oriented classification method to extract urban ecological land. In the preparation process classification rules set introduced NDVI, NDWI, length/width, image brightness and nearest neighbor classification method to classify. Setting 2008 ZY-1 02 B star data as a reference, monitoring the 2008-2013 Urban ecological land use changes in Yuhua district in Shijiazhuang City.The results show: the traditional classification of high image information extraction method has some limitations. But supervised classification accuracy is better than unsupervised classification, the traditional classification methods for information extraction is suitable for large area, such as water. Object-oriented classification is better than the traditional classification method. Extracting features high spectral images, geometric and texture characteristics. Not only the classification accuracy was significantly higher than the traditional classification methods, but also takes full advantage of high image provided information to facilitate monitoring of urban ecological land.The GF-1 interpretation classification rule set is obtained by comparing the results of interpretation. When the segmentation scale is 80, NDWI greater than 0.11, the brightness is less than 207, you can extract the water; when splitting scale of 25, NDVI greater than 0.05 can extract vegetation; Use nearest neighbor classification method, select the sample point may be bare, roads and construction sites to distinguish. When the image is set to determine the classification rules, the paper choose Yuhua District, Shijiazhuang City as a typical area. Contrast 2008 ZY-02 B star scores imaging and 2013 GF-1 image interpretation results. It found in its urbanization process, ecological land decrease significantly.