Research on vision and trajectory planning system for tomato picking robots

Abstract

This article comprehensively considered the working characteristics of the robot and the way of growing tomatoes. The traditional planting method was transformed at the beginning. And we designed a four-axis robotic arm on this basis. This research combined with the Kinect V2 camera to build an experimental platform to conduct research on vision algorithms and planning algorithms. Then the Kinect V2 camera-manipulator-tomato three-dimensional coordinate transformation model was established through visual calibration on ROS. The construction of the control system was completed, and finally the picking test was carried out. The experimental results showed that the method was convenient and efficient, and completed the picking tomatoes well. It has important reference value for the future research of unmanned tomato planting in the harvesting system in the greenhouse.