论文部分内容阅读
以南宁市为研究区,利用手持GPS实地调查取样,获得甘蔗及其它作物训练样本区。通过分析甘蔗等作物年度内光谱特征变化规律,依据不同作物物候历不同的原理,利用多时相HJ-1CCD卫星数据,采用监督分类、决策树分类和逐步剔除相结合的遥感方法,提取研究区甘蔗种植信息。野外验证和分类精度分析表明,无论从空间分布趋势还是面积数据来看,其信息精确度均较高,分类精度达到92.3%。由此可见,环境减灾卫星数据结合分类方法能基本满足制糖企业对甘蔗种植面积提取精度的要求,可以在制糖企业中推广应用。
Taking Nanning city as the research area, using hand-held GPS field survey sampling, get sugarcane and other crop training sample area. Based on the different principles of different climatic dates, using the satellite data of HJ-1CCD from multiple phases, the remote sensing method combining supervised classification, decision tree classification and step-by-step elimination was used to analyze sugarcane and other crops’ annual variation of spectral characteristics. information. Field validation and classification accuracy analysis showed that the accuracy of information was high, with a classification accuracy of 92.3%, both in terms of spatial distribution trend and area data. Thus, the combination of environmental disaster reduction satellite data classification methods can basically meet the requirements of sugar extraction of sugar cane plantation area extraction accuracy, which can promote the use of sugar companies.