• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Big Earth Data >A generalized supervised classification scheme to produce provincial wetland inventory maps: an application of Google Earth Engine for big geo data processing
【24h】

A generalized supervised classification scheme to produce provincial wetland inventory maps: an application of Google Earth Engine for big geo data processing

机译：生成省级湿地清单地图的广义监督分类方案：Google Earth Engine在大地理数据处理中的应用

获取原文
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Wetlands are important natural resources due to their numerousecological services. Consequently, identifying their locations andextents is imperative. The stability, repeatability, cost-effectiveness,multi-scale coverage, and proper spatial resolution imagery of satel-litesprovideavaluableopportunityfortheiruseinvariouslarge-scaleapplications, such as provincial wetland mapping. To do so, it isrequired to (1) process and classify big geo data (i.e. a large amountof satellite datasets) in a time- and computationally-efficientapproach and (2) collect a large amount of field samples. In thisstudy, Google Earth Engine (GEE) and machine learning algorithmswere utilized to process thousands of remote sensing images andproduce provincial wetland inventory maps of the three Canadianprovinces of Manitoba, Quebec, and Newfoundland and Labrador(NL). Additionally, using GEE, a generalized supervised classificationmethod is proposed to produce a regional wetland map from a largearea (e.g., a province) when lacking field samples. In fact, using thefield data from only Manitoba and assuming that all wetlands inCanada have similar characteristics, the wetland maps were gener-ated for the other two provinces. The overall classification accuraciesfor Manitoba,Quebec, and NLwere84%,78%,and 82%, respectively,indicating the high potential of the proposed method for aidingprovincial wetland inventory systems.
机译：湿地由于其众多的生态服务而成为重要的自然资源。因此，必须确定其位置和范围。卫星-lites的稳定性，可重复性，成本效益，多尺度覆盖和适当的空间分辨率图像为其在各种大型应用（例如省级湿地测绘）中的使用提供了宝贵的机会。为此，需要（1）在时间和计算效率高的方法中处理和分类大的地理数据（即大量的卫星数据集），以及（2）收集大量的现场样本。在这项研究中，Google Earth Engine（GEE）和机器学习算法被用于处理成千上万的遥感图像，并绘制出曼尼托巴，魁北克，纽芬兰和拉布拉多（NL）三个加拿大省的湿地清单图。另外，使用GEE，提出了一种通用的监督分类方法，以在缺少野外样本时从大区域（例如，一个省）生成区域湿地图。实际上，仅使用曼尼托巴省的实地数据，并假设加拿大的所有湿地都具有相似的特征，便为其他两个省生成了湿地图。曼尼托巴省，魁北克省和自然保护区的总体分类准确度分别为84％，78％和82％，表明所提出的方法可用于省级湿地清单系统。

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