摘要
为污水灌溉区水稻生理健康及重金属胁迫诊断提供理论依据,以吉林省长春市西湖污水灌溉区水稻为研究对象,通过ASD3光谱仪测量水稻光谱,SPAD-502便携式叶绿素仪获取水稻叶绿素含量,采用支持向量机和PROSPECT模型,构建了长春市西湖污水灌溉区水稻叶绿素含量的高光谱估算模型。结果表明:支持向量机建立的水稻叶绿素含量估算模型的测试精度较好,一阶微分变换有助于提高估算模型获取水稻叶绿素含量的精度;水稻叶绿素高光谱估算模型的最大相关系数为0.8563,最小均方根误差(RMSE)为9.5106。估算模型可在精细农业管理中快速、有效地估算水稻的叶绿素含量。
The spectrum and chlorophyll content of rice irrigated with sewage were measured by a ASD3 spectrometer and SPAD-502 portable chlorophyll meter,and then the hyperspectral estimation model for chlorophyll content of rice irrigated with sewage was built based on support vector machine and PROSPECT model to provide the theoretical basis for rice physical health and diagnosis of heavy metals stress in Xihu sewage irrigation area,Changchun City,Jilin Province.The results showed that the test precision of the estimation model for determining rice chlorophyll content built by a support vector machine is better,and first-order differential conversion can help to improve the test precision of rice chlorophyll content determined by the estimation model.The maximum correlation coefficient and minimum root-mean-square error of the estimation model is 0.8563 and 9.5106 respectively.The estimation model can estimate rice chlorophyll content in management of precision agriculture rapidly and effectively.
出处
《贵州农业科学》
CAS
北大核心
2014年第8期237-239,248,共4页
Guizhou Agricultural Sciences
基金
江西省数字国土重点实验室开放研究基金项目"矿区尾矿重金属污染的多源遥感协同监测研究"(DLLJ201305)
地理空间信息湖南省工程实验室开放研究基金项目"尾矿区土壤重金属含量的高光谱遥感反演研究"(2013GSIJJ002)
农业部农业信息技术重点实验室开放课题"尾矿区农田重金属污染的高光谱遥感监测研究"(2013006)
西安科技大学博士科研启动项目"矿区尾矿重金属污染的多源遥感协同监测研究"(2012QDJ036)
关键词
污灌区
水稻
叶绿素含量
估算模型
高光谱
sewage irrigation area
rice
chlorophyll content
estimation model
hyperspectrum
