基于Landsat影像的雄安新区2014～2018年土地利用变化检测

Land Use Change Detection Based on Landsat Image in Xiongan New Area(2014～2018)

雄安新区是继深圳经济特区和上海浦东新区之后又一意义重大的新区,该地区的土地利用变化研究具有重要意义。但目前对于雄安新区初建阶段的土地利用变化的研究比较有限。本研究基于2014、2018年的雄安新区的Landsat影像对土地利用类型先进行面向对象的监督分类,再对比分类结果进行变化检测。基于支持向量机和k近邻方法两种分类方式得到的结果表明:雄安新区的建成区的增长百分比约为17%~21%,农田比例降低4%~5%,闲置农田约有20%的比例转化为建成区。结合夜光遥感影像提升了分类精度,同时利用面向对象的方法解决了同谱异物和同物异谱的问题。本研究的区域选择具有时效性,紧贴国家千年大计,为相关决策者提供了宏观信息支持,为雄安新区其他的相关研究提供了参考。

The Xiongan New District is another important state-level area after Shenzhen Special Economic Zone and Shanghai Pudong New District.it is necessary to evaluate the construction speed and quality of Xiongan New District and to serve the development strategy in line with local conditions.However,the research on land use change in the initial stage of Xiongan New District is limited.Based on the Landsat image of Xiongan New District in April 2014 and April 2018,this study applied the support vector machine and kth-nearest neighbour methods to detect the land use change of the land in Xiongan New District.The results obtained by the two classification methods indicate that the growth percentage of the built-up area of Xiongan New district is about 17%~21%.The proportion of agricultural land has reduced by 4%~5%,while about 20%of the idle farmland has been converted into built-up areas.This study can serve as reference for other related research and future policy formulation in Xiongan New District.