扎龙湿地景观的遥感信息提取

Remote Sensing Information Extraction of Zhalong Wetland Landspace

文章以扎龙自然保护区为研究区域,利用2014年与2020年扎龙湿地遥感影像作为数据源,对遥感影像进行处理,获取研究区内各种不同地表类型的训练样点,提取训练样本,得到各自的训练数据,利用神经网络分类方法对整个扎龙湿地地表覆盖类型进行分类,得到扎龙湿地景观类型分类图。通过2014年与2020年2a的影像对比,得出6a来扎龙湿地的各部分面积变化,分析6a间各部分景观面积变化的影响因素,为湿地生态系统的研究及保护工作提供数据基础。结果表明:神经网络分类结果较为精确,2014-2020年6a间,扎龙自然保护区内草地面积增加明显;明水、有水湿地面积略有增加;而农田、盐碱地、以及无水湿地面积减少,其中无水湿地面积减退较多,湿地整体情况稳中向好。

In Zhalong nature reserve as the research area,using 2014 and 2020 in Zhalong wetland remote sensing image as data sources,the remote sensing image processing,access to research in a variety of different types of training sample,extract the training sample,get their own training data,using the neural network classification method for the whole of Zhalong wetland surface coverage types are classified,Zhalong wetland landscape type classification figure.Through the image comparison between 2014 and 2020,the area change of each part of Zhalong Wetland in the past five years is obtained,which provides a data basis for wetland ecosystem research and protection work.The results showed that the neural network classification results were accurate.During six years from 2014 to 2020,the grassland area in Zhalong Nature Reserve increased significantly.The area of wetlands with clear water and water increased slightly.The area of farmland,saline-alkali land and anhydrous wetland decreased,among which the area of anhydrous wetland decreased more,and the overall situation of wetland was stable and improving.