Real-time reconstruction of orange tree model based on obstacle avoidance

Based on obstacle avoidance of orange fruit harvesting robot, the real-time fruit tree model reconstruction is studied. Modular modeling method is adopted for the reconstruction, which extracts a standard module from the orange fruit tree by configuration study and then gets the fruit tree model after combining the module. Through the comparison between truncated cone module and cylinder module, step cylinder section is chose as the standard module. In order to supply the tree model in time, polygonal approximation is used in the standard module construction for the data reduction. The experiment for middle complexity tree model proved, that this method can realize real-time reconstruction of orange tree model, and the model is precise enough for obstacle avoidance.