面向对象的泰安市城市附属绿地信息提取

Information Extraction of Urban Affiliated Greenland in Tai'an City Based on an Object-oriented Method

城市附属绿地广泛分布在居民区和道路等与人类工作生活最息息相关的区域。基于面向对象的分类方法,选取泰安市高分辨率遥感影像Quick Bird,对城市附属绿地信息进行提取,并与监督分类结果进行比较评价。结果表明,面向对象分类的总体精度达到了84.15%,Kappa系数为0.816 1,明显高于监督分类。泰安市居民区、单位和道路绿化程度总体较低,需要加强居民区和单位附属绿地建设,并提高道路绿化水平。

Urban affiliated greenland can be found between roads, in residence and other areas closely related to human life and work. Information of urban affiliated greenland in Tai＇an City was extracted from the high spatial resolution remote sensing image Quick Bird based on an ob- ject-oriented approach. The average precision of classification is 84.15%, and the Kappa coeffi- cient is 0. 816 1. This shows that this method is significantly improved in classification compared with the supervised classification. This study indicates that the green percentage of Tai＇an is fairly low and more work needs to be done to improve it.