摘要
建立数学模型来分析树叶的形状、分类及质量。首先,通过抽取树叶的重要参数,利用聚类分析进行树叶分类,并用BP神经网络进一步验证模型;其次,利用三维投影分析得到树叶相互遮挡对树叶形状的影响,并结合光照在树内的分布以及改进的Logistic模型,得到在树内不同位置的树叶大小规律。将树的分枝类型分为两类,通过统计学分析得出树木的不同分枝类型与树叶形状之间的显著相关性。最后,提出了分别基于树冠体积计算和树木生长规律的两种树叶质量估算模型,并通过比较得到了最优解。
In this paper,mathematical models are built to analyze the shape,classification and mass estimation of leaves.Firstly,several key parameters of leaves are picked out to do cluster analysis.After that a neural network is built to test the classification.Secondly,to analyze the relation between leaf shape and overlapping,a 3Dprojection model is set up and the light distribution in the tree is figured out with revised Logistic model.Furthermore,tree branching patterns are divided into two types to find out the significant correlation between branching patterns and leave shapes.Finally,two models are put forward to estimate the mass of leaves,where the first one is the volumetric model and the second one is based on the growth mechanism of the trees.By detailed comparison of the two models,the optimal solution of this problem is obtained.
出处
《数学建模及其应用》
2012年第2期50-59,共10页
Mathematical Modeling and Its Applications
