面向数据共享的城镇化信息快速提取技术方法与应用——以江苏盐城为例

Study on Rapid Extraction Technique of Data Sharing Oriented Urbanization Information: A Case in Yancheng City,Jiangsu Province,China

随着世界经济的发展,全球城镇化进程正在以前所未有的速度进行。城镇化带来的耕地占用、生态环境破坏、水资源短缺等问题日益受到关注。快速获取和共享城镇变化信息对于人地关系研究及可持续发展意义重大。借助遥感手段监测城镇变化信息,已经普遍得到应用,目前,多光谱遥感数据是最经济和可行的遥感监测数据源。本文基于多光谱卫星数据,有效利用城镇地区特有的地物光谱特征,建立不同类别城镇信息的提取模式和技术流程。这些可以通过多光谱遥感数据提取的信息包括城镇建成区、水体和绿地等。选择我国东部沿海区域内城镇化进程明显的江苏盐城主城区作为试验点,基于地物光谱特征分析,依据本文建立的城镇化信息提取流程,采用阈值分析和二值化处理方法,实现了盐城市亭湖区主要城镇区2000、2002、2006年三期城镇信息快速提取,使在更大范围内,促进此类信息的快速获取和共享,具有现实意义。

With the development of world economy,land change and land cover in urban areas tend to change drastically over very short time due to rapid urbanization.Urbanization pushes the development of regional economy,also enhances the integrated power of the country.While at the same time,urbanization also brings negative influences for ecological environment regionally or globally,such as cultivated land occupation,ecological degradation,water shortage,etc.How to acquire urbanization information quickly and share these data in time is a key problem.With the development on Earth observation technique,remote sensing has been one of the powerful methods for urbanization monitoring.Multi-spectrum remotely sensed images are widely used because they are economical and accessible easily,such as Landsat TM/ETM＋ and so on.Based on multi-spectrum images,the paper studies various land use information extraction technique flow according to urban objects＇ spectrum features.Based on the designed research technique flow,4 kinds of urbanization information are extracted from remotely sensed images,i.e.,the built-up urban area,water area,green belts and thermal islands.Yancheng City is selected as a case for the urban built-up area extraction.The built up area in Yancheng City was extracted automatically by means of designed method from three Landsat images in 2000,2002 and 2006.Firstly,Yancheng City＇s spectrum features are analyzed.Secondly,some urbanization information is automatically calculated through models of NDBI,NDVI and NDWI.Thirdly,binary processing method is used in the experiment based on the experiential threshold values of NDBI,NDVI and NDWI.Finally,the built up area in the main urban zone can be calculated and analyzed.This experiment shows that it is feasible and effective to extract some urbanization information from multi-spectrum images through remote sensing technologies,such as water area,green belts,city built up area,etc.This will enhance urbanization information rapid extraction and acquirement in more cities.Through data sharing,this information will be applied in the process of social and economic development in China.