Design of Delta Robot Using Image Processing for Product Sorting Process

Abstract

This paper describes the Delta robot design and control method that combines image processing to classify products with the advantage of high speed and accuracy. The Delta robot has three degrees of freedom of paralleled kinetic structure and closed-loop structure. There are three plastic bars, with each bar, one end is connected to the mobile rack via a parallelogram configuration. With such a geometric structure, the end platform is always oriented and only moves forward along the 3 axes of XYZ. Objects to be classified are identified by the camera and then sent to the PC with the coordinates, color, and size of the object. The PC that receives the data will calculate through the dynamic equation combined with image processing to find the coordinates of the real-time object, then send the information to the PLC to control the robot to pick and drop it into the classification box. The system has been installed and tested in reality. The percentage of finding colors and picking the right objects is 97.8%. The Delta Robot control software is programmed with both Auto and Manual modes, simulating the industrial product sorting process.