Object Tracking and Following Robot Using Color-Based Vision Recognition for Library Environment

Abstract

In the current era of Industrial Revolution 4.0, many robots are expected to co-exist and assist humans, in different environments. It would be a common scene that years from now, service robots follow humans to help us in our everyday work. Researchers investigated various methods and techniques for a mobile robot to track and follow objects or humans. Our proposed method was to use a color-recognition scheme and decide on the motion of the robot based on pixel value (which represents area) of the identified object. The robot will move towards the object if the pixel value is lower than a threshold value (an indication of a distant object), move away if the pixel value is higher than a threshold value (the object too close) or stay still if the pixel value is within these two values. Implementation on a library environment shows the success of the algorithm, with only around average 4cm deviation from the 10 trials tested. The significance of the findings is that the robot is able to maintain its distance from the object, thus avoiding from colliding with the object or being too far from the object.