【作者】 张俊宁； 李民赞； 张小超；

【Author】 Zhang Junning1, Li Minzan2, Zhang Xiaochao1 (1. Chinese Academy of Agricultural Mechanization Sciences, Beijing, 100083; 2. China Agricultural University, Beijing, 10083)

【机构】 中国农业机械化科学研究院； 中国农业大学；

【摘要】 土壤数据提供了土壤中静态的理化属性信息,而作物光谱数据提供了作物动态生长信息,两者一静一动,取长补短,有机的结合提供了相对完整的农田作物生长环境信息,为实现精准的作物产量估计提供了可靠数据保障。本文以农田土壤信息和作物冠层高光谱信息为基础,综合分析它们与冬小麦产量的相互关系,并采用非线性建模方法开展土壤和作物生长信息融合的冬小麦产量估测研究,建立冬小麦产量的高可靠性预测模型,并进行模型精度比较分析。结果表明,基于信息融合的神经网络模型在全部回归模型中达到最优预测效果,体现出土壤空间分布信息与作物生长信息融合的冬小麦产量预测方法的优越性和重要性。更多还原

【Abstract】 Soil data shows static physical and chemical attributes of field soil, and spectral data of crop provides dynamic information of crop growth, so the effective combination of these two kinds of data above can supply a relatively integral crop information in field which will provide guarantee for realizing crop yield estimation accurately. Information of soil and wheat canopy spectrum were considered as essential data to roundly analyze the relationship with winter wheat yield, and were carried out research about crop yield estimation based on information fusion in this paper. The Artificial Neural Network (ANN) prediction models were established and showed optimal prediction results, and this will reveal the benefits and importance of modeling based on the integration of soil data and crop growth data further.更多还原