论文部分内容阅读
针对无人机遥感油菜估产问题,提出基于全约束混合像元分析方法的油菜估产模型。针对油菜冠层构成实际特征,分析了不同地面端元构建方式对油菜无人机影像光谱分解的影响,在此基础上,分别在油菜开花期和油菜荚果期建立了影像丰度数据和地面实测产量数据的关联。实验分析表明,所提出的混合光谱油菜产量估算方法具有较好的效果。油菜荚果期和开花期估产模型的相关系数R~2分别为0.765 2和0.721 2,综合两个时期的估产模型相关系数R~2为0.814,说明在油菜生长的不同时期,目标端元丰度与油菜产量具有较强的相关性,证明了该模型具有较高的精度和较强的稳定性。
In order to solve the problem of remote sensing of rapeseed by UAV, a rapeseed estimation model based on full constrained mixed pixel analysis is proposed. According to the actual characteristics of canola canopy composition, the effect of different ground terminal building methods on spectral decomposition of rapeseed UAV image was analyzed. Based on this, the image abundance data and ground measured data were established respectively at flowering and rapeseed rapeseed Production data association. Experimental results show that the proposed hybrid spectral rapeseed yield estimation method has a good effect. The correlation coefficient R ~ 2 of rapeseed pod and flowering period estimated yield models were 0.765 2 and 0.721 2, respectively. The correlation coefficient R ~ 2 of the estimated yield model during the two periods was 0.814, indicating that at different stages of rapeseed growth, And rapeseed yield has a strong correlation, proves that the model has high accuracy and strong stability.