Simulation and experiment of tomato pollen particles release and motion characteristics based on optical flow target tracking method

Abstract

Due to the reduction of the amount of natural pollinators, it is necessary to study the method and mechanism of greenhouse tomato mechanical vibration supplementary pollination. The release process and motion characteristics of pollen particles are very important for precision tomato supplementary pollination. In this paper, the pollen particles release and motion simulation model based on discrete element and hydrodynamics coupling method is established. The pollen particles position and trajectories detection algorithm based on optical flow target tracking method is designed. Compared with the detection results of high speed photography system, the relative error of pollen distribution range angle and pollen motion velocity of the simulation model is 7.94% and 8.87%, which shows the rationality and reliability of the established simulation model. Through the comprehensive analysis of the high speed photography observation results and simulation results, it can be concluded that the trend of pollen particles trajectories are vertical downward with random direction change, and greater vibration amplitude cause greater distribution range angle of pollen particles. There is no obvious trend of velocity change during the release process of each single pollen particle, and the relationship between vibration amplitude and average velocity of pollen particles is not obvious. The research results and the established model of this paper provide a mechanism basis for tomato mechanical vibration supplementary pollination and lay a foundation for the optimization of mechanical pollination method.