山西农业大学学报（自然科学版）2017，Vol.37Issue(7)：510-516,7.
黄土高原露天煤矿区面向对象地貌自动分类方法研究
Automated classification of landform elements using object-based image analysis in large surface coal mine of loess plateau
孙瑞 1马萧 2潘雪婷 3王洪光1
作者信息
- 1. 江苏省地质测绘院,江苏南京211102
- 2. 北京岩土工程勘察院,北京100083
- 3. 北京市地质工程设计研究院,北京101500
- 折叠
摘要
Abstract
[Objective]This study aimed to detect the landform change from loess plateau to artificial landscape in large surface mining area using object-based classification method.[Methods]Object-based classification method has been widely applied in remote sensing imagery and large scale landform classification.The Digital Elevation Model (DEM) of 10m resolution in 1990 and 2010 were selected as data source.The object-based classification rules were established based on the topographic parameters of slope,profile curvature,plane curvature,and relative elevation,which were generated from DEM.[Results]The results showed that:According to the break point from the change rate of local variance,the optimal segmentation scale was 14 for natural loess landform and 11 for artificial dump landform.The landform classification rules based on membership function overcame setting specific threshold in most of the automatic classification method,which could provide consistency landform element.The objected-oriented image analysis considered the characteristics like shape,geometric and neighborhood proved to be more readily adaptable for mine and loess plateau as the overall accuracy of loess plateau and mine could reach to 74.47％ and 81.31％.According the objectbased landform classification from dump area,it was found that the landform element of convex shoulder and convex backslope accounted for the largest area of the whole dump,which was 30.24％ and 20.32％ respectively.[Conclusion]The object-based landform classification method with effective and robust rules could provide the basis for the artificial landform reconstruction and land reclamation in the dump area.关键词
面向对象/地貌分类/黄土高原/排土场Key words
Object-based image/Landform classification/Loess plateau/Mine dump分类
农业科技引用本文复制引用
孙瑞,马萧,潘雪婷,王洪光..黄土高原露天煤矿区面向对象地貌自动分类方法研究[J].山西农业大学学报（自然科学版）,2017,37(7):510-516,7.
