Three- Dimensional Modeling and Visualization of Rice Root System Based on the Improved Dual-Scale Automaton and L-system
Le Yang,Panpan Wu,Mingfu Liao,Jun Peng,Zizong Tang,Huibin Long,Huanhuan Zhang,Xiaoyun Yu
DOI: https://doi.org/10.1016/j.compag.2022.106823
IF: 8.3
2022-01-01
Computers and Electronics in Agriculture
Abstract:In agronomy sciences and researches, root system architecture (RSA) of Oryza sativa L. (rice) is often used to reflect the spatial configuration of rice mots. But, because soil has its own opacity, the cognition and expression of the spatial morphology and structure of rice roots have become the bottleneck in the in-depth study. The three-dimensional (3D) modeling and visualization can be used to help further study and recognize the morphological, structural and functional traits of rice roots. To clarify the rules governing the structure and distribution of the rice root system, and so to understand the relationship between RSA and the functionality of rice mots, a method combined the improved dual-scale automaton with Lindenmayer-system (L-system) for the three-dimensional modeling and visualization of the rice root system is hereby proposed. The basic parameters of the rice root growth were firstly measured by the destructive detection, and various appropriate growth functions were selected according to the rice root growth rule and development so as to control the rice root growth rate in the modeling. In the dual-scale automaton modeling, the identifiers of micro-states and macro-states were redefined with numbers and characters, and hence the improved dual-scale automaton model was used to describe the growth process of rice roots, which was then combined with L-system grammar. According to the morphological structure characteristics of the rice root growth, the direction and radius of the rice root growth were constrained, and a single three-dimensional morphological model was thus constructed. Then, the visualization of the rice root growth was realized by MATLAB. By calculating the total rice root length, total rice root surface area, total rice root volume, maximum rice root width, maximum rice root depth and other indicators of the model, as well as analyzing and comparing the experimental measurement data, it was finally found that the average accuracy of the simulation of the total rice root length, total rice root surface area and total rice root volume was 94.27%, 93.68% and 92.35%, respectively. Besides, the maximum rice root width and maximum rice root depth had strong correlation with the corresponding measured data. So, the results showed that the model had good simulation effects on the rice morphological structure. Again, based on this, the relationship between the rice root structure and its functions was further analyzed by quantifying the structural parameters such as the rice root length fraction, solidity, convex hull volume, rice root volume density and rice root surface area density, etc. This model laid a foundation for the coupling of both the morphological structure model of the rice root growth as well as the model of physiological and ecological factors of next stage, hoping to provide a reference for the 3D modeling and visualization research of the root growth of other crops.