| 摘要: |
| 该文采用日本ALOS卫星AVNIR多光谱影像，选取山地丘陵草丛、山地丘陵灌木草丛、山地丘陵 疏林草丛3类为研究对象，运用决策树和ArcGIS空间分析方法，提取的草地类型总体精度为84.03%， Kappa系数为0.755 6，草地类型信息的获取为喀斯特山区草地资源监测、估产、生态畜牧业发展与规划等 研究和政府决策提供参考。 |
| 关键词: 喀斯特山区 草地类型 ALOS影像 决策树分类 |
| DOI：10.7621/cjarrp.1005-9121.20130612 |
| 分类号: |
| 基金项目:基金项目:国家重点基础研究计划（973计划）（2012CB723202）;贵州省国际科技合作计划（黔科合外G字［2012］7022号）;贵州 省优秀青年科技人才培养对象专项资金（黔科合人字（2009）18号）;贵州省科学技术厅与贵州师范大学联合基金（黔科合J字LKS ［2011］40号）;贵州师范大学创新基金项目资助（研［2012］14） |
|
| RESEARCH OF THE GRASSLAND TYPES EXTRACTION BY ALOS IMAGE IN KARST MOUNTAINOUS AREA |
|
Yang Jinli1, Zhou Zhongfa1, Zhao Zhenglong2, Zhou Zeying3
|
|
1.Institute of South China Karst， Guizhou Normal University， Guiyang 550001/The State Key Laboratory Incubation Base for Karst Mountain Ecology Environment of Guizhou Province， Guiyang 550001;2.The planning office of agricultural resources Buyi and Miao Nationalities of Qiannan Autonomous Prefecture in Guizhou， Duyun 558000;3.The Forage workstation Buyi and Miao Nationalities of Qiannan Autonomous Prefecture in Guizhou， Duyun 558000
|
| Abstract: |
| Based on the AVNIR multispectral image from Japan's ALOS satellite， this paper selected mountains and hills grassland， mountains and hills shrub grassland， and mountains and hills open forest grassland as the ob- jects， used the decision-making tree and ArcGIS method to extract the grassland types. The results showed that the overall accuracy was 84.03%， Kappa coefficient was 0.755 6. It could provide an important reference for grassland resources monitoring， yield estimation， the development research of ecological animal husbandry， and the government decision-making in Karst mountainous area. |
| Key words: karst grassland type ALOS image decision tree |
