Construction of Robotics and Application of the Optical-Flow Algorithm in Determining Robot Motions

Abstract

This article presents the research results in the application of image processing in determining the position, direction, and moving speed of different objects when they move in the free space in the field of view of measuring cameras. The research includes developing an algorithm to detect, identify, and locate objects and an algorithm to calculate the movement direction and instantaneous velocity of the object. Two robots with measurement systems were designed for testing, one carries the investigating object, and the other the camera. These robots can communicate with a computer system using a tele-wireless system. A program was also built for capturing images and analyzing the state of the model automatically. Experimental results show that the position, angle, and velocity of different objects can be captured well. The average error in determining the direction of movement is an average of 1.25°, and the error of the moving speed is less than 0.5 m/s. The research results provide a potential tool for designing robots for highly effective detection. The algorithm and the measurement system are simple and inexpensive but highly effective, and they can be used in the initial process of determining an object. A detailed description of the algorithm, robot system, and testing will be presented in this study.

Keywords

• optical flow
• robot
• computer vision
• controlling robot
• image processing