| 摘要: |
| 由于干旱区农牧交错地带湿地面积小且地类混杂,混合像元现象严重,使得该区湿地信息的自动提取难度较大。针对湿地遥感信息提取的特点和难点,采用NDVI(normalized different vegetation index)阈值法提取水体,应用支持向量机模型(SVM)提取去除水体后的湿地信息,并以修正线性光谱混合分解模型分解的草甸分量,进一步提取高盖度、中盖度和低盖度草甸信息。试验验证结果表明:提取结果的总体精度为88%,Kappa系数为0.83。该方法可为其他光谱特征混杂地区湿地遥感信息的提取提供参考。 |
| 关键词: 湿地 线性光谱混合分解模型 遥感 农牧交错地带 植被覆盖度 |
| DOI：10.11841/j.issn.1007-4333.2011.03.024 |
| 投稿时间：2010-09-19 |
| 基金项目:北京林业大学研究生科技创新专项计划项目资助(BLYJ201103); 国家"十一五"科技支撑计划(2006BAD23B05) |
|
| Wetland information extraction based on improvedlinear spectral mixture model |
|
WU Jian, PENG Dao-li
|
| (The Key Laboratory for Silviculture and Conservation of Ministry of Education,Beijing Forestry University,Beijing 100083,China) |
| Abstract: |
| Automatic extraction of wetland information in farming-pastoral regions is very difficult,because land use types are complex and etland area is small which lead to erious mixed pixel phenomenon.Aming at the characteristics and difficulties of wetland remote sensing information extraction in these regions,the NDVI (Normalized Different Vegetation Index) threshold method was used to extract water body first.Afterwards,the support vector machine(SVM) was selected to extract wetland information which has removed water.Finally,the meadow component which was decomposed by linear spectral mixture model was amended by the high-resolution satellite image to further extract high coverage,medium coverage and low coverage meadow.The experimental results showed that the overall accuracy was 88% and the Kappa coefficient was 0.83.This method can provide a reference of remote sensing information extraction of wetland for other regions with mixed spectral characteristics. |
| Key words: wetland linear spectral mixture model remote sensing farming-pastoral area vegetation coverage |
