Simulation of apple tree branch growth using an improved point-cloud-boundary-constraint model

The apple is a valuable commercial crop. Morphological information on and visual simulation of the apple plant are critical for precise and scientific field management during its growth period. However, the complex morphological structure of apple trees make it difficult to simulate the growth process. According to the pipeline model theory, branches play a key role in the growth process of apple trees. In this study, an improved semi-Markov-based point cloud boundary constraint is proposed to simulate the growth process of apple tree branches. To improve simulation accuracy, the point cloud data of the dormant, bud, strip-pulling, and leaf curtain stable stages of the apple tree were continuously collected, and the point cloud boundaries of various growing stages were extracted to obtain constrains for the growth process of branches. The results revealed that the improved semi-Markov-based point cloud boundary constraint model can achieve highly accurate simulation of the growth process of apple tree branches, with an average branch reconstruction rate of test samples in various growth periods and an average skeleton point coincidence rate of 93.0% and 90.8%, respectively.