基于时序哨兵1号数据的黑龙港流域农作物类别精细提取研究

Research on Fine Extraction of Crop Categories in Heilonggang Basin Based on Time Series Sentinel-1 Data

摘要: 针对光学遥感数据容易受云、雨等天气影响的不足，本文以黑龙港典型区域为实验区，基于随机森林分类方法探索了采用哨兵1号时序雷达数据为数据源进行农作物精细分类。结果表明：1) Senti-nel-1雷达数据受云、雨等天气状况的影响较小，因此能够构建更加完整的特征曲线以反映农作物生长信息；2) 利用Sentienl-1数据提取VH、VV极化时序数据和纹理特征并构建多种分类方案，其中多时相VH、VV双极化时序数据分类精度最高，纹理特征数据的加入并没有明显提高分类精度。

Abstract: In view of the vulnerability of optical remote sensing data is impacted by weather conditions such as clouds and rain, this article takes the typical area of Heilonggang as the experimental area and explores the use of Sentinel-1 time series radar data as the data source for crop fine classification based on random forest classification method. The results show that: 1) Sentinel-1 radar data is less affected by weather conditions such as clouds and rain, so it can construct more complete feature curves to reflect crop growth information, which has great application value in extracting crop planting structure in the experimental area; 2) Using Sentienl-1 data to extract VH and VV polarization time series data and texture features and construct a variety of classification schemes, the results show that the multi-phase VH and VV dual polarization time series data have the highest classification accuracy, and the addition of texture feature data does not significantly improve the classification accuracy.