基于SVM的高分时序冬小麦种植区提取研究

Study on Extraction of Winter Wheat Planting Area based on SVM

冬小麦作为我国重要的粮食作物,准确获取其空间分布情况,对农业生产管理及农情监测有重要意义。以河南省商丘市为例,利用覆盖冬小麦完整生育期的GF-1数据,计算归一化植被指数(Normalized Difference Vegetation Index,NDVI)、增强植被指数(Enhanced Vegetation Index,EVI)时间序列,结合关键生育期影像,构建不同特征量组合数据集,利用支持向量机方法进行冬小麦提取。同时采用主成分分析法对数据进行降维处理,尝试通过压缩特征集数据量来提高冬小麦提取效率。研究结果表明:EVI时序数据较NDVI能更好地描述作物的物候,提取精度皆高于NDVI,其中EVI时序数据与关键生育期影像组合提取精度最高,达到97.67%。结果表明,降维后数据并未对提取精度造成显著影响,达到压缩数据量保持提取精度的目的,为大区域作物提取提供参考价值。

As an important food crop in China,winter wheat has accurate access to its spatial distribution and is of great significance for agricultural production management and agricultural monitoring. Taking Shangqiu City of Henan Province as an example,using the GF-1 data covering the whole growth period of winter wheat,the time series of Normalized Difference Vegetation Index(NDVI)and Enhanced Vegetation Index(EVI)was calculated. During the growth period,the data sets of different feature quantities were constructed,and the winter wheat was extracted by the support vector machine method. The results show that the EVI time series data can better describe the phenology of crops than NDVI,and the extraction accuracy is higher than NDVI. The EVI time series data and the key growth period image combination extraction precision is the highest,reaching97.67%. At the same time,the principal component analysis method is used to reduce the dimensionality of the data,and try to improve the extraction efficiency of winter wheat by compressing the data volume of the feature set. The results show that the data after dimension reduction does not have a significant impact on the extraction accuracy,and the purpose of maintaining the accuracy of the compressed data is to provide reference value for large-area crop extraction.