Application of Object-oriented Classification in Extraction of Impervious Degree of Urban Surface

In this paper,based on the multi-resolution image segmentation and object-oriented image classification,the information of the impervious degree of urban surface in Nanjing was extracted by using the objects relationship at the different scales.The results indicated that the imperious degree in Nanjing City was mainly concentrated between 50% and 99%,and the area of higher degree was extensively distributed in Jianye District,Xiaguan District and Gulou District,while the lower was mainly in Xuanwu District.The method of object-oriented image classification could realize the quick extraction of the impervious degree of urban surface,and the output was stability and reliability.