论文部分内容阅读
在分析以数字图像处理为测定法确定林带疏透度误差来源的基础上,对其中的随机误差进行研究,结果表明:林带整体疏透度的机误方差小于其冠部和干部的较大者;所研究的各类型林带各部位疏透度随机误差均遵从正态分布;林带整体疏透度随机误差的分布与树种和带内配置无关;北京杨、双阳快杨和其它类杂交杨的矩形或品字形配置林带各自冠部与干部疏透度机误方差之间无显著差异,而乡土杨林带干部的显著大于冠部的。本文还分别各类型林带的各部位确定了由林带疏透度测定值估计其总体实际值的随机误差限,并讨论了在测定林带疏透度过程中据该误差限计算样本量和划定测定范围的应用意义。最后总结提出:以增加测定同一林带不同样段像片数限定随机误差,通过模型计算订正疏透度测定值中的投影误差和影缩误差确定林带疏透度是建立完善的“数字图像处理法确定林带疏透度”新方法的可行途径。
Based on the analysis of the digital image processing method to determine the source of the error of forest belt density, the random error was studied. The results showed that the error variance of the overall permeability of forest belt was smaller than that of crown and cadre . The random errors of all parts of the shelter belts in each type of forest belonged to the normal distribution. The distribution of random errors in the overall permeability of the shelter belts had nothing to do with the tree species and in-band configuration. Rectangles of Beijing Yang, Shuangyang express and other hybrid poplars Or the shape of the forest configuration with the crown and cadres, respectively, there is no significant difference between the machine variance variance, while the native Yanglin cadre was significantly larger than the crown. In this paper, we also determine the random error limits of the overall actual value estimated by the measured values of the forest belt’s dazzle in all parts of each type of forest belt, and discuss the calculation of the sample size and demarcation according to the error limit Scope of the application of significance. In the end, we conclude that it is a perfect method to establish the perfect digital image processing method by determining the random error by increasing the number of images in different sections of the same belt and by calculating the projection error and shadowing error in the measured value of the measured transparency. Determination of forest permeability "a new way of feasible ways.