基于图像处理技术的树木疏透度确定方法研究

Research on the method of determining tree porosity based on image recognition technology

如何确定树木的疏透度一直是研究的热点,应用现有的方法确定疏透度时多需要人工干预。文中提出了确定单株树木疏透度的新方法,该方法基于图像处理技术对树木的照片进行识别即可确定树木的疏透度。选取九里香六年苗、八年苗(Murraya exotica)和豆瓣黄杨(Buxus sinica)三株与真实树木外观相似的树苗作为实验材料,利用数码相机拍摄不同风速下的树木照片,再应用图像处理技术中阈值法以及分水岭算法对照片进行处理,从像素的层面通过计算获得单株树木疏透度。将运用文中方法所得到的树木疏透度与前人运用图像处理工具Photoshop所获得的疏透度相比较,结果表明:三株树木在各风速状态下的疏透度差异在2%~8%之间。文中提出的基于图像处理的树木疏透度确定方法精度比较高,而且该方法具备不需要人工干预、操作简单的优势;同时发现三株树木的疏透度均随风速增加而增大。

How to determine the degree of transparency of trees has always been a hot research topic.The application of existing methods to determine the degree of transparency often requires manual intervention.This paper proposes a new method to determine the degree of transparency of individual trees.This method is based on image processing technology to identify the photos of trees to determine the degree of transparency of trees.In this paper,three saplings of Jiulixiang six-year seedlings,eight-year seedlings and Douban boxwood trees with similar appearance to real trees were selected as experimental materials.The digital camera was used to take photos of the trees under different wind speeds,and then the threshold method in the image processing technology was applied.And the watershed algorithm is used to process the photos,and the transparency of individual trees is obtained from the pixel level through calculation.Comparing the transparency of the trees obtained by the method in this paper with the transparency obtained by the predecessors using the image processing tool Photoshop,the results show that the difference in the permeability of the three trees under each wind speed is 2%~8%.The method for determining the degree of transparency of trees based on image processing proposed in this paper has relatively high accuracy,and this method has the advantages of not requiring manual intervention and simple operation.At the same time,it is found that the degree of transparency of the three trees all increase with the increase of wind speed.