摘要
以乌鲁木齐市为例,采用高分辨率卫星QuickBird数据为主要信息源,应用面向对象的多尺度分割技术,对城市绿地信息的提取方法进行了研究。利用专题层分割影像,结合成员隶属度函数,实现城市绿地信息的自动提取。研究表明使用该方法提取信息速度快、准确率高、成本低,可以实现城市绿地信息的快速而准确的提取。
Monitoring urban green space is one of the major environmental applications in remote sensing today. As the main data sources for urban green space high-resolution imagery provides a good basis for recognizing and monitoring small-scale structure changes. Going far beyond the methodical limits of pixebbased and manual interpretation approaches, multi-resolution image segmentation and object-oriented image analysis approaches are used for extracting information from satellite remote sensing data. This paper discusses about the extraction of urban green space from a QUICKBIRD image using this new technique. It allows the segmentation of an image into highly homogeneous image objects in any chosen resolution and generation of a network of image objects. The process does not classify single pixel but rather image object. Not only spectral information but also spatial, physical and contextual characteristics of image objects are used for classification. The green space has been effectively extracted from the QuickBird image by using Fuzzy Classification Method that based on rules. The result of green space information extraction is promising and precision of classification is higher.
出处
《福建林业科技》
北大核心
2006年第1期40-44,共5页
Journal of Fujian Forestry Science and Technology
