Colored Object Sorting using 5 DoF Robot Arm based Artificial Neural Network (ANN) Method

摘要

Automation process is very important to help any industry activity especially to sorting an object. In this research, we build 5 DoF of Robot Arm based on Arduino microcontroller to pick and place colored object intended sorting task in real-time. The algorithm of interfaces and image processing to detect the colored object based on Python 2.7 with OpenCV library. From image processing system, the detected colored object based on Red-Green-Blue (RGB) identifier has a coordinate. The coordinate data is collected to build the inverse kinematic model and trained by using Artificial Neural Network (ANN) method to control Robot Arm. Finally, the algorithm implemented to Robot Arm to do pick and place a colored object show a good accuracy rate. The advantages of this Colored Object Sorting system are Open Source and Low Cost.