节点文献
基于SPOT5高分辨率影像的城市新增建设用地监测研究
Monitoring of New Construction Land Using High Resolution SPOT5 Imagery
【作者】 侯洪升;
【导师】 赵庚星;
【作者基本信息】 山东农业大学 , 农业推广, 2009, 硕士
【摘要】 人多地少是我国的基本国情。近年来,随着经济的快速发展和城镇化水平的不断提高,建设用地迅速增加,而农用地和未利用地则越来越少,土地资源和社会经济发展的矛盾日益突出,在城市地区表现尤为明显:城市建设用地多以粗放的土地利用方式快速增长,并大量占用耕地等农用土地,严重影响着土地的合理利用和城市的可持续发展。为严格土地管理,节约集约用地,切实提高土地利用效率,国土资源管理部门在全国广泛地开展了国土资源大调查、土地利用动态遥感监测等工作。基于此,本文以泰安市区为例,采用先进的“3S”技术和地面调查测量等技术手段,对城市新增建设用地进行动态监测研究,以便为国土资源管理工作提供准确翔实的依据。本文在国内外遥感监测研究方法的基础上,以2006年和2007年两时相的SPOT5全色影像和多光谱影像为数据源,制作成研究区的遥感正射影像图,采用计算机自动提取技术和人机交互目视解译方法提取变化信息,以此实现对城市建设用地、耕地等各类用地的遥感动态监测。在遥感正射影像图制作过程中,系统地阐述了影像的配准、融合和正射纠正等过程的理论与方法:以2006年SPOT5全色影像为参考,在图像匹配方法的基础上,对其它相关数据源进行配准;之后对常用的四种图像融合方法进行了试验对比和结果评价,选取效果较好的PCA变换法作为主要的融合方法;通过PCI和ERDAS软件平台对融合处理后的图像进行正射纠正,提高图像判别精度。在此基础上,采用图像差值法、光谱特征变异法等计算机自动提取技术,辅助人机交互目视解译的方式来提取变化图斑;之后,参照相关图件和统计资料对变化信息进行外业调查核实,并对调查结果做后期内业处理,确定变化图斑的类型、位置等属性信息,建立城市新增建设用地数据库,输出成果图并统计汇总监测结果。通过对监测结果的研究分析可知:(1)新增建设用地监测精度符合要求:遥感正射影像图误差在1个像元之内,监测图斑的属性和面积等误差均不超过10%,并通过外业验证已消除;(2)监测期间,研究区新增建设用地的面积为2413.39亩,变化图斑数为102块,且岱岳区变化程度明显大于泰山区;(3)新增建设用地类型为工矿仓储用地、住宅用地和商服用地等五种,主要来源是占用耕地和其它非耕农用地;(4)从时间上看,研究区建设用地年度变化较为合理,并且有增加的趋势;从空间上看,新增建设用地多在城市边缘地带的城乡结合部,且沿主要道路、工矿及居住等用地分布;(5)新增建设用地的主要驱动力为社会经济发展、城乡建设、人口及政策规划等。最后,根据变化原因和相关问题,提出了四点可行性措施,为土地资源管理提供参考性建议。
【Abstract】 One of the basic national conditions of China is much Population with less land. In recent years, with the rapid development of economy and the improvement of urbanization, construction sites are increasing rapidly while farmland and unused land are less and less, so the contradiction has manifested itself between land resources and the development of social economy, especially in urban areas: the construction sites grow rapidly in the form of extensive land use, and lots of farmlands are occupied, as a consequence, the rational utilization of land and the sustainable development of the city are seriously affected. For the purpose of strict land management, using land more economically and intensively as well as improving the efficiency of land use, various regions widely develop land resources surveys. Based on this, the paper chooses Taian city proper as an example to carry out dynamic monitoring study of new construction sites by“3S”and the survey measurement technologies, so as to provide accurate informative basis for land resources management.Based on the research methods of both domestic and foreign remote sensing monitoring, the study produced remote sensing orthophoto maps of research area with 2006 and 2007 SPOT5 panchromatic images and multi-spectral images data, and then used the computer automatic extraction technology and man-machine interactive visual interpretation method to extract change information, so as to realize remote sensing dynamic monitoring of urban construction sites, arable land and other kinds of land. During the process of producing orthophoto maps, the paper systematically expounded theories and methods of the images registration, images fusion and ortho rectification: based on image matching method, it took 2006 SPOT5 panchromatic images as the reference to registering other related data; then carried on experimental comparative analysis and effect appraise to four common methods of image fusion, and chose PCA transformation method as the main method; in order to improve image distinguishing accuracy, the images were put into ortho rectification. On the base of image processing, change spots were extracted with man-machine interactive visual interpretation method, which was helped by computer automatic extract technology such as image margin, spectrum variability and so on; then according to relevant datum and statistical data to carry out field investigation and late-stage indoor work, so that the types, places and other attributes information of change spots can be defined, moreover, to establish the database of urban new construction sites, output maps and count statistic monitoring results.The results of research indicate that: (1) The monitoring accuracy accords with the requirements: the error of orthophotomap is within one pixel limits,also,the attribute and area errors are not more than 10%, and these errors have be avoided through field confirmation; (2) During the monitoring period, new construction sites cover an area of 2413.39 mu, the number of change spots is 102, and the change extent of District de Daiyue significantly greater than District de Taishan’s; (3) New construction sites have five types, including industrial warehousing land, residential land, business land and so on; (4) From the time, the annual change of construction sites is reasonable, and has an increasing trend; from the space, new construction sites are located in the integration of urban and rural or suburban areas, and mostly distribute along the main roads, factories and residential land use, etc; (5) The main driving forces of new construction sites are economic and social development, urban construction, demographic factor and planning policy. Finally, according to the change reasons and related problems, the paper puts forward four feasible measures, to make referential suggestions for land resources management.
【Key words】 remote sensing; dynamic monitoring; new construction sites; change information; research area;
- 【网络出版投稿人】 山东农业大学 【网络出版年期】2012年 02期
- 【分类号】F293.2;P237
- 【被引频次】5
- 【下载频次】344