Color-Based Image Processing for Autonomous Human Following Trolley Robot Navigation with Camera Vision

Abstract

The rapid advancements in the field of robotics have spurred intensive research, particularly in the industrial sector, aiming to develop robots that can assist in simplifying daily human tasks. One emerging area of research involves the design of a cargo-carrying robot trolley. This trolley robot has the capability to follow a person carrying items by recognizing the color of the clothes worn by that person through image processing. The objective of this research is to facilitate the transportation of goods, especially in airport environments, by enabling the robot to identify and follow human objects with a minimum distance of 30 centimeters and a maximum distance of over 3 meters. The design system of this robot trolley utilizes a camera sensor to detect the object to be followed through image processing using OpenCV on the Microsoft Visual Studio 2012 platform. The image processing results in PWM values sent to the Arduino to drive DC motors. Additionally, ultrasonic sensors are employed to restrict the robot's movement in its surroundings, preventing collisions. The robot's speed can adjust according to the walking speed of a person. If the robot is moving too fast, it will be stopped by the ultrasonic sensor when the distance between the robot and the person being followed is less than 30 cm, avoiding collisions between the robot and the person.