基于无人机图像分形特征的油松受灾级别判定
发布时间：2019-03-21 14:49
【摘要】：利用无人机采集油松样地图像,提取图像中的单株样本树图像,计算单株样本树图像的多个纹理特征值,对纹理特征值进行灾害分级,与地面基于失叶率调查的灾害分级进行比对,探索能准确描述油松受灾情况的无人机图像纹理特征。实验结果表明,受灾油松图像的三种分形特征,即分形维数、缝隙量及维数升降因子能较好地反映油松的失叶率状况,可作为油松受灾级别的图像判定特征,同时上述分形特征也适用于整块油松样地的受灾级别判定。
[Abstract]:An unmanned aerial vehicle (UAV) was used to collect the image of Pinus tabulaeformis sample plot, extract the image of a single sample tree, calculate the multi-texture eigenvalues of the image of a single sample tree, and classify the feature values of the texture. Compared with the disaster classification based on the investigation of leaf loss rate on the ground, the texture features of UAV images which can accurately describe the disaster situation of Pinus tabulaeformis are explored. The experimental results show that the three fractal features of the image of Pinus tabulaeformis (Pinus tabulaeformis), that is, fractal dimension, slit volume and dimension rise and fall factor, can well reflect the leaf loss rate of Pinus tabulaeformis, and can be used as the image decision feature of the disaster level of Pinus tabulaeformis. At the same time, the fractal features mentioned above can also be used to determine the disaster level of the whole plot of Pinus tabulaeformis.
【作者单位】： 北京林业大学信息学院;北京林业大学林学院;
【基金】：林业公益性行业科研专项资助项目(201404401)
【分类号】：S763.7;TP391.41
[Abstract]:An unmanned aerial vehicle (UAV) was used to collect the image of Pinus tabulaeformis sample plot, extract the image of a single sample tree, calculate the multi-texture eigenvalues of the image of a single sample tree, and classify the feature values of the texture. Compared with the disaster classification based on the investigation of leaf loss rate on the ground, the texture features of UAV images which can accurately describe the disaster situation of Pinus tabulaeformis are explored. The experimental results show that the three fractal features of the image of Pinus tabulaeformis (Pinus tabulaeformis), that is, fractal dimension, slit volume and dimension rise and fall factor, can well reflect the leaf loss rate of Pinus tabulaeformis, and can be used as the image decision feature of the disaster level of Pinus tabulaeformis. At the same time, the fractal features mentioned above can also be used to determine the disaster level of the whole plot of Pinus tabulaeformis.
【作者单位】： 北京林业大学信息学院;北京林业大学林学院;
【基金】：林业公益性行业科研专项资助项目(201404401)
【分类号】：S763.7;TP391.41
【相似文献】
相关期刊论文 前6条
1 于海鹏;刘一星;刘镇波;;基于图像纹理特征的木材树种识别[J];林业科学;2007年04期
2 张还;何春霞;侯人鸾;于e，
本文编号：2445051
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2445051.html
