中文摘要：

农村居民地遥感信息获取对于监测农村居民地时空变化、服务＂三农＂和国土资源管理具有重要意义。以Rapid Eye卫星影像数据为数据源,江西省泰和县为研究区,利用最大似然法进行居民地等土地覆盖类别的分类提取与精度评价及分析。结果表明：分类总体精度达到84.33%,其中农村居民地的制图精度和用户精度分别为76.01%和82.28%,与第二次全国土地调查中的居民地数据对比,其一致性达到71.0%。结合实地验证,对本分类精度的误差原因进行分析。本研究表明利用5 m分辨率的Rapid Eye影像进行县级农村居民地监测是可行的,可为后续同类研究提供技术参考。

英文摘要：

The monitoring of the spatial and temporal changes of rural residents is of great significance in serving ＂three rural issues＂,land and resources management based on rural remote sensing information. In order to monitor the residential area especially the rural area by using remote sensing technology,the authors chose Taihe County in Jiangxi Province as the study area. On the basis of the Rapid Eye satellite images after ortho- rectification with spatial resolution of 5 meters,the residential land classification extraction and accuracy evaluation were carried out using the maximum likelihood method. The results showed that the overall classification accuracy reached 84. 33%.The producer＇s accuracy and user＇s accuracy of the residential land were 76. 01% and 82. 28% respectively,and the consistency reached 71. 0% in comparison with the residential land data of Taihe County in the second national land survey,and the possible causes of errors were also analyzed. The research results show that the resident land monitoring of rural counties using Rapid Eye images of 5 meters resolution is feasible in that it can provide reference for future similar studies.