面向对象的农村用地信息遥感提取技术应用——以宁夏西吉为例

发布时间：2018-03-19 19:00

本文选题：面向对象　切入点：多尺度　出处：《城市发展研究》2017年09期 　论文类型：期刊论文

【摘要】：在我国城镇化建设过程中,农村空心化成为一个不可避免的难题。农村用地信息提取是空心村整治的基础。研究选取宁夏西吉县典型生态移民村作为空心村信息提取的研究对象,结合高分辨率DOM遥感影像,采用面向对象的分类方法,在多尺度分割基础上提取不同地物类型信息,并与土地调查数据进行对比,结果表明总体精度在80%以上。在此基础上提取空心村房屋信息并与宅基地数据进行对比,精度在75%以上。研究表明面向对象的空心村信息遥感提取技术可行,精度高,可为空心村整治提供基础信息和数据保障。
[Abstract]:In the process of urbanization in China, Rural hollowing becomes an inevitable problem. The extraction of rural land information is the basis for the renovation of hollow villages. The typical ecological immigrant villages in Xiji County of Ningxia are selected as the research objects of information extraction of hollow villages. Combined with high resolution DOM remote sensing image, the object oriented classification method was used to extract the information of different ground objects on the basis of multi-scale segmentation, and compared with the land survey data. The results show that the total precision is more than 80%. Based on this, the information of hollow village houses is extracted and compared with the data of homestead, the precision is more than 75%. The research shows that the technology of remote sensing extraction of hollow village information based on object-oriented is feasible and has high precision. Can provide basic information and data guarantee for hollow village renovation.
【作者单位】：宁夏回族自治区国土资源地理信息中心;
【基金】：国家“十二五”科技支撑计划课题资助项目(2014BAL01B02)
【分类号】：S127

【相似文献】