摘要
运用基于对象的影像分析方法,对秦王川地区1994、2001、2011、2013年同时相的4期Landsat 5 TM、Landsat 8 OLI卫星影像数据进行影像分割,综合运用影像对象的光谱、纹理、形状和邻域特征,构建了分类规则集,实现了各期土地利用/土地覆被专题制图.结果表明4期影像的分类结果精度均满足制图要求与动态监控,准确地反映了过去20 a来秦王川地区土地利用/土地覆被的状况.
Object-based image analysis were conducted to create image objects of Qinwangchuan area of Lanzhou,by image segmentation of four scenes of images acquired in 1994,2001,2011,2013 derived from Landsat 5 TM and Landsat 8 OLI satellites.The integrated multiple characteristics associated with spectrum,texture,shape,and neighborhood features of image object were employed to construct rule sets for classification of land use and land cover,so as to realize thematic mapping of these four periods.The results showed that accuracy of image classification are qualified for thematic mapping and dynamic monitoring in Qinwangchuan area of Lanzhou during the last 20 years.
出处
《兰州大学学报（自然科学版）》
CAS
CSCD
北大核心
2016年第1期37-42,共6页
Journal of Lanzhou University(Natural Sciences)
基金
国家自然科学基金项目(41271360)
国家基础科学人才培养基金科研训练与科研能力提高项目(2012-2016)
