基于多载荷无人机遥感的大豆地上鲜生物量反演

发布时间：2018-02-08 21:17

本文关键词： 无人机遥感载荷生物量植被指数大豆　出处：《大豆科学》2017年01期 　论文类型：期刊论文

【摘要】：以多载荷无人机获取数据和地面实测的数据为基础,将大豆生殖生长期分段建模,采用植被指数和光谱参数相结合再加上农学参数株高,通过最小二乘法建立多元线性回规模型的方法,来估算大豆开花期和结荚期的鲜生物量,采用高光谱植被指数法估算大豆鼓粒期和成熟期的鲜生物量。结果表明:在大豆开花期和结荚期内,采用混合法构建生物量反演模型利用交叉验证法,验证结果的R~2和RMSE分别为0.714和0.393;在大豆鼓粒期和成熟期内,采用高光谱植被指数法构建生物量反演模型,利用交叉验证法,验证结果的R~2和RMSE分别为0.697和0.386;大豆开花结荚期构建的模型和鼓粒成熟期构建的模型都有比较高的精度和可靠性,利用这两种模型完成了高光谱影像鲜生物量的遥感空间制图,能反映当地当时大豆的真实长势情况。
[Abstract]:Based on the data obtained from the multi-payload UAV and the data measured on the ground, the soybean reproductive growth stage was modeled in sections, and the vegetation index and spectral parameters were combined with the agronomic parameter plant height. To estimate the fresh biomass of soybean in flowering and pod-setting stages, a multivariate linear regression scale model was established by using the least square method. The hyperspectral vegetation index method was used to estimate the fresh biomass of soybean at grain filling stage and mature stage. The results showed that the hybrid method was used to construct the biomass inversion model during the flowering and pod-setting stages of soybean and the cross-validation method was used. The results showed that RX2 and RMSE were 0.714 and 0.393, respectively. The hyperspectral vegetation index method was used to construct the biomass inversion model, and the cross-validation method was used. The results showed that RX2 and RMSE were 0.697 and 0.386, respectively. The models constructed at flowering and pod-setting stage of soybean and the models constructed at grain filling stage had high accuracy and reliability. The spatial mapping of fresh biomass in hyperspectral image was completed by using these two models. Can reflect the real growth of the local soybean at that time.
【作者单位】：河南理工大学测绘与国土信息工程学院;北京农业信息技术研究中心;南京农业大学国家大豆改良中心;
【基金】：北京市自然科学基金(4141001) 北京市农林科学院科技创新能力建设项目(KJCX20140417) 国家自然科学基金(41271345,41471351)
【分类号】：S565.1;S127

【相似文献】