首页> 外文会议>Asian Conference on Remote Sensing;Asian Space Conference >Land Cover Land Use Classification Using Multi-sensor,Multi-temporal Satellite Data; Mekong Delta, South Vietnam
【24h】

Land Cover Land Use Classification Using Multi-sensor,Multi-temporal Satellite Data; Mekong Delta, South Vietnam

机译:使用多传感器,多时间卫星数据进行土地覆盖土地使用分类; 湄公河三角洲,南越南

获取原文

摘要

Remote sensing data, Landsat-ETM, JERS-OPS/SAR data in Mekong River delta (Lat 10°30'-10°45', Lon 106°45'-107° 00') were processed for Land-cover Land-use classification. The delta, a flood plane of Mekong River, comprises of a river network formed due to huge freshwater supply from the Mekong basin for millions years. On the other hand during dry period, low river discharge tends to back flow of ocean water due to tides (3-4 m) along the drainage system, causing salt water intrusion into upstream results salt water contamination into the environment. Therefore, the land-cover consists of salt marshes with dense Mangrove forest including other kind of brackish water flora. JERS-OPS images were processed to discriminate the Land-cover classes. Not like in the other areas, wetlands consist of both freshwater and brackish water where, Agriculture, aquaculture, salt mashes. Capability of discriminating those classes are difficult using optical sensor data, hence multi sensor and multi-temporal approaches have been made to overcome the situation to improve the results. The processing technique adopted in this study is to identify major classes which contain similar spectral characteristics and mask out to segregate major classes. This will enhance the image seperability to perform secondary classification to each segregated major classes and avoid mixing among classes. Near Infra Red (NIR) band of optical sensors allowed to separate three major classes such as water, wetlands and vegetation. However, vegetation represents Mangrove forests, Plantations, paddy and other agricultural crops while the wetlands consist of salt marshes with sparse mangroves and other brackish water flora, mud flats, and aquaculture practices a challenge for image processing. Multi-temporal and multi sensor data allows to overcome the challenge by means of changing nature of land use pattern. Methodology and the results of this work are discussed in this paper in detail.
机译:遥感数据,Landsat-ETM,Mekong River Delta(Lat 10°30'-10°45',Lon 106°45'-107°00')的陆地覆盖土地使用分类。湄公河洪泛飞机的三角洲包括由湄公河盆地较大的淡水供应占据了一百万年的河流网络。另一方面,在干燥期间,低河流放电由于沿着排水系统潮汐(3-4米)而导致海水流动,导致盐水侵入上游结果盐水污染到环境中。因此,陆地覆盖包括盐沼,浓密的红树林森林,包括其他类型的咸水植物群。处理图像以区分陆地覆盖类。不像在其他地区,湿地包括淡水和咸水,农业,水产养殖,盐泥。使用光学传感器数据难以判断这些类的能力,因此已经进行了多个传感器和多时间方法来克服这种情况以改善结果。本研究采用的处理技术是识别含有类似的光谱特性和掩模以隔离主要类的主要类别。这将增强对每个分离的主要类执行次要分类的图像单个性,并避免在类之间混合。靠近红外线(NIR)光学传感器带允许分离三个主要类,如水,湿地和植被。然而,植被代表红树林林,种植园,稻谷和其他农作物,而湿地由盐沼与稀疏红树林和其他咸水植物群,泥舱和水产养殖做出挑战的盐沼。多时间和多传感器数据允许通过改变土地利用模式的性质来克服挑战。本文详细讨论了该工作的方法和结果。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 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