基于不同滴灌方式的苹果幼树根系生长的三维可视化模拟

为研究不同滴灌方式下大田苹果幼树的根系生长行为,提出一种用Otsu方法对根系图像进行分割后统计分析根系空间分布的方法,利用图像处理库通过Python语言编程实现了获取空间分布统计矩阵,以Lynch的根构型模型为结构原型,结合功能模型建立不同滴灌方式下的大田苹果幼树根系生长行为结构功能模型,最后利用OpenAlea对该模型进行三维可视化模拟。以大田苹果幼树为试验对象进行验证,同时依据形态特征的相似度以及仿真数据与实测数据的相关性两个参考指标将模拟结果与实物进行对比,验证了所采用算法的有效性。结果表明,Otsu算法可有效分割苹果幼树根系图像数据,结合常规功能结构参数和空间分布统计矩阵建立的大田苹果幼树根系结构功能模型可科学地模拟不同滴灌方式对苹果幼树根系空间拓扑的影响,大田苹果幼树根系结构功能模型模拟可以可视化不同滴灌方式下的根系生长空间变化行为。可视化模拟结果表明,分根区交替灌溉的苹果幼树根系在根区两侧分布比较均匀,根密度较大,有利于提高根系对水肥的吸收利用效率。研究结果可为分根区交替灌溉在果园的应用提供参考。

Modeling and visualization of three-dimensional young apple tree root growth under different modes of drip irrigation

Roots absorb not only nutrients and water, but also chemically signal the state of soil nutrient and propagate the signal in soils to control physiological activities of above-ground plant organs such as stomata opening/closing, leaf blade expansion and fruit development. Therefore establishing and maintaining a healthy root system in the ground for normal development of stem, leaf and fruit is critical in crop production. But because most plant roots and growth environments are not visible, the spatial configuration and distribution of plant roots are difficult to map out. Plant visualization simulation provides quick visual effect for studying plant root growth behavior. Hence model simulation of three-dimensional plant root conformation and distribution in soils and computer visual expression are vital for the interpretation of the absorption and utilization of water and nutrient by root. This study simulated young apple tree root growth behavior under field conditions of different modes of drip irrigation. The study proposed the Otsu method to statistically analyze root distribution in space after root image partitioning. By using image processing library to obtain spatial distribution statistics matrix in Python language programming, a structure-function model was established for young apple tree root distribution under different modes of field drip irrigation. The Lynch root structure model was used as a prototype of the combined root structure-function model. Finally, three-dimensional visualization of young apple tree root growth was simulated by using the OpenAlea model. Using field young apple trees for verification, the results showed that the Otsu algorithm effectively segmented young apple tree root image data. The young apple tree root structure-functional model was established by combining general function-structure parameters and spatial distribution statistics matrix scientifically, and the model could scientifically simulate the spatial topology of young apple tree roots under different modes of field drip irrigation. The structure-functional model of young apple tree root was suitable for visualizing the changing behavior of root growth in space under different drip irrigation modes in the field. The simulated visualization results suggested that the distribution of young apple tree roots was more even under alternate irrigation of partial root zone (i.e., the two sides of the root zone). Root density was also higher under alternate irrigation, which improved water and fertilizer uptake and use efficiency. The study provided additional theoretical reference for the application of alternate irrigation of partial root zones in orchard fields.