3种冠型分维数求算法在杨树无性系中的应用

A Comparison of Methods for Estimating Fractal Dimension of Poplar Crown

对盒维数法、双表面积法及贝氏法等3种树冠分维数估算方法在杨树无性系上的应用结果进行了比较。结果表明:双表面积法操作复杂且工作量大;贝氏法虽然操作简单但是精度太低;盒维数法具有操作简便、快速、重演性高、能充分反映树冠内叶面积的空间分布模式等优点,但目前仅能估算2维平面的分维数。相对而言,盒维数法应该是求算树冠分维数最理想的方法。树冠的分维数与树冠形状、树冠大小相关性不大,而与树冠内叶面积的多少密切相关,一般地,树冠内叶面积越多,树冠的分维数越大。

Three methods （Box-counting method, The two-surface method and Berezovskava method ） which were applied to estimate fractal dimension of poplar crown were compared in this study. The two-surface method was of heavy work and complexity. Berezovskava method was proved to be unsuitable for estimating fractal dimension of poplar crown due to its inaccuracy. In spite of limitation to 2-dimension, it is safe to say that Box-counting method was the best way to estimate fractal dimension of poplar crown owing to its easy operation, rapidity, high repetition, good stand for crown structure and foliage distribution. There was no link between fractal dimension of crown and crown shape, crown size, but a significant correlation was found between fractal dimension and foliage. Generally, the larger the foliage is, the bigger the fractal dimension is.