利用二维图像评价“Y”形苹果幼树的分形特征

摘要

【Objective】Assessing quantitatively the fruit tree architecture provides theoretical value and guiding significance to conduct training and pruning on fruit tree.【Method】In this research,based on the fractal theory and the method of box-counting dimension,we investigated the fractal characteristics of young apple trees trained to Y-shaped configuration among five cultivars by two-dimension images,and regarded the fractal dimension as an indicator of assessing the fruit tree architecture.【Result】Fractal dimensions of young apple trees trained to Y-shaped configuration differed from their cultivars;For the same cultivar,the fractal dimensions differed between front view and side view;For five cultivars,their fractal dimensions of front view varied from 1.697 to 1.554;Side view,from 1.699 to 1.563;For the same cultivar,the fractal dimension values of side view were higher than those of front view;The fractal dimension was positively corrected with the length and number of shoots;By comparing the fractal dimension of five cultivars of apple tree,the comprehensive evaluation showed this:Nagafu 2〉Yuhua Zaofu〉Fuhong Zaoga〉Yanfu 6〉Li Gala.【Conclusion】Fractal dimension can reflect objectively the spatial distribution and ability of shoots to utilize space,thus it can be the quantitative indicator of tree architecture,and can guide the training and pruning on young apple tree trained to Y-shaped configuration.%【目的】定量化评价果树的树体结构，为果树的合理整形修剪提供指导。【方法】基于分形理论，采用计盒维数的方法，将果树二维图像分形维数作为评价树体结构的指标，研究了长富2号、玉华早富、富红早嘎、烟富6号和丽嘎啦等5个苹果品种“Y”形幼树二维图像的分形特征。【结果】不同品种苹果幼树二维图像的分形维数有差异，同一品种果树正、侧视图像的分形维数存在差异，正视图的分形维数介于1.697～1.554，侧视图的分形维数介于1.699～1.563，同一品种侧视图的分形维数大于正视图，分形维数与果树枝条长度、数量呈正相关关系，5个苹果品种幼树二维图像的分形维数表现为长富2号〉玉华早富〉富红早嘎〉烟富6号〉丽嘎啦。【结论】分形维数能客观地反映果树干枝的空间分布及空间利用程度，可作为评价苹果幼树树体结构的定量化指标，用于指导苹果幼树的整形修剪。