Study On Spatial Location System Of Target Fruit Based On Point Laser Targeting Measurement

Because of the low level of mechanization and intellectualization in China's agricultural production,the research of agricultural harvesting robot has received more and more attention.And because all picking operations are dependent on the identification and orientation of the visual system as well as the research of vision system is the most important part of the picking robot.Therefore,the research on the vision technology of the picking robot is one of the most concerned topics in the field of agricultural automation.This paper studies a method of the spatial location system of the cherry tomato based on point laser targeting measurement.Aimed at providing a basis for picking robot to perform picking action,which need obtain the space coordinates of cherry tomato.The main contents and methods are as follows:Firstly,the vision system platform based on point laser targeting control is built.The camera's internal and external parameters are obtained by calibrating the 2D template.And using Pre-D-H parameters method to build geometry structure of the 6 DOF series robot and obtain direct kinematics equation.Secondly,we acquire the image features that extract by CogPMAlignTool and the estimation of the coordinates which at highest,and then complete the manipulator target based on visual servo.Then the accurate depth of target fruit grain is measured by laser range finder.Based on location algorithm,spatial coordinate of tomato is calculated.Test result shows that the location error in the X and Y direction is less than 6mm and the location error in the Z direction is less than 10 mm.Finally,the picking test is carried out in the greenhouse.The vision system is installed on the tomato picking platform.The rate of successfully positioning fruit is 83%,the result shows that the vision positioning system based on point laser targeting measurement meets the requirements of the actual picking.