Design and simulation robotic arm with computer vision for inspection process

Abstract

Different companies compete for control of the size of the commercial markets and to offer quality products. Companies are keen to examine the products and ensure that they are free of any defects or distortions before distributing them in markets in order to preserve the confidence of customers, but manual inspection is expensive and takes a lot of time. Investors tended to use modern technologies to implement the examination process. In this paper, an approach based on the association of robots with a computer vision system is proposed. A robot arm with 4 degrees of freedom is designed by SOLIDWORKS software that takes the cans into the conveyor belt then passes it to the examination room so that an image of the product is taken via camera attached to the computer and the image is processed by the LABVIEW program. The model was simulated using MATLAB, and Arduino microcontroller has been used for controlling the processes perform by the prototype. When a defective product passes a conveyor belt, the system changes the path to remove the product from the production line. The simulation and experimental results proved that the prototype is capable of grasping cans, then detecting the cans finally, taking the defective ones out of the production line. By using this technology in the product sorting process, the productivity will be increased and the quality will be enhanced in a short time. An accuracy of 96% has been accomplished by the proposed system, where of 50 samples only two samples haven’t detected.