首页> 外文会议>Asian conference on remote sensingACRS >MANGROVE FOREST EXTENT MAPPING IN SIARGAO ISLAND, SURIGAO DEL NORTE, PHILIPPINES USING SUPPORT VECTOR MACHINE (SVM) CLASSIFICATION
【24h】

MANGROVE FOREST EXTENT MAPPING IN SIARGAO ISLAND, SURIGAO DEL NORTE, PHILIPPINES USING SUPPORT VECTOR MACHINE (SVM) CLASSIFICATION

机译:红树林森林范围映射在瑞典德德尔·北部,菲律宾使用支持向量机(SVM)分类

获取原文

摘要

Accurate spatial information of mangrove forest extent is essential for both land-use planning and natural resources management. In this study, we assessed the spatial distribution of mangrove forest in Siargao Island, Surigao del Norte for the years 2000 and 2010 using publicly available Landsat data. Landsat images from 1994-2000 and 2006-2010 were Landsat calibrated and atmospherically corrected using the FLA ASH algorithm. Support Vector Machine (SVM) classification was then used to classify the images to 4 classes (i.e., Mangroves, Terrestrial non-MangToves, Cloud Cover, and Shadows). Our results indicate that the total area of mangrove forest cover was approximately 62.4 km~2 and 79.9 km~2 for years 2000 and 2010, respectively, with overall accuracy of 87% (2000) and 96% (2010) and kappa coefficients 0.789 (2000) and 0.939 (2010). Our findings differ from previous mangrove extent estimates (73.6 km~2) done in the Philippines by Long and Giri (2011) and Long et al. (2013), which showed no difference in mangrove forest extent between 2000 and 2010. In general, our results indicate that there is an overall increase in mangrove extent on most parts of the island especially on pre-existing mangrove stands. However, it can also be observed that some stands significantly decreased in cover in areas near built-up regions. The results of this study are important contributions to ecological and natural resources assessment, conservation and management studies.
机译:准确的红树林范围空间信息对于土地利用规划和自然资源管理至关重要。在这项研究中,我们评估了2000年和2010年苏格罗岛的红树林森林的空间分布,使用公开的Landsat数据。 1994-2000和2006 - 2010年的Landsat图像是使用FLA灰分算法校准和大气校正的Landsat。然后支持支持向量机(SVM)分类,将图像分类为4类(即,红树林,地面非漫步,云盖和阴影)。我们的结果表明,2000年和2010年,红树林林覆盖的总面积约为62.4公里〜2和79.9 km〜2,整体准确性为87%(2000)和96%(2010)和Kappa系数0.789( 2000)和0.939(2010)。我们的调查结果与以往的红树林范围估计(73.6 km〜2)不同于菲律宾(2011)和龙等人。 (2013),在2000年和2010年期间,红树林森林范围内没有差异。一般来说,我们的结果表明,在岛上大部分地区的大多数地区都有整体增加了红树林范围,特别是在预先存在的红树林中。然而,也可以观察到,在建筑物附近的区域覆盖范围内有些代理显着下降。本研究的结果是对生态和自然资源评估,保护和管理研究的重要贡献。

著录项

相似文献

  • 外文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号