| As an import part of the forest resources, woodland, is also one of indicative factors of Echo-environment statement. It plays an important role in air purification, natural vaccination, oxygen, and other aspects climate regulation. In recent years, with its unique characteristics of comprehensive,macro,fast and so on.Satellite Remote Sensing Technology was widely applied to agriculture,forestry,meteorology,oceanography,environmental protection and other fields. At the same time the resolution of satellite images have greatly improved along with the rapid development of Satellite Remote Sensing Technology.Images can reflect abundant and meticulous land features which can provide strong support for the forest resources effectively extract and real-time dynamic monitoring.In this paper,Yongqing country which is located in the center of Langfang city in Hebei province and abundant forest resources was selected as study area. GF-1(8-m multispectral and 2-m panchromatic) was used to discuss the object-oriented methodology of identification and extraction of woodland information.Firstly,the pretreatment to the expression data,including orthorectification,image fusion,cutting stitching,geometric correction and other processes. Secondly,the optimal segmentation scale parameters were got by considering the global features of the image;and then woodland information fusion was carried out based on fuzzy logic classification algoriths,and extraction was achieved combine images and textures,the spectral characteristics.Lastly,the extracted accuracy was compared with the traditional pixel-based classification method.It is concluded that the extraction Kappa coefficient of woodland information by the object-oriented classification was 0.89,and the overall accuracy was 94.4%,at the same time traditional method of classification reached 0.66 and the overall accuracy was 80%;The result indicated that object-oriented approach has great potential to be used in woodland information extraction.
|
PDF Full Text Request