A binocular vision system for object tracking and distance perception based on optical convergence

• Main Authors: Ke, Gaston Anthony N., Ong, Ivan Lawrence D., Sy, Rosvic C.
• Format: text
• Language: English
• Published: Animo Repository 2006
• Subjects: Computer vision; Robot vision; Image processing; Artificial intelligence; Pattern recognition systems
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/7040
• Institution: De La Salle University

The continuous expanding applications of robotics leads to a trend, that is to create robots with at least the basic functions of a human being in order to accomplish tasks with quality equal to or even greater than that done by a human. One basic and probably the most important element in an effective robotic system is its vision. Our system applies the principle of convergence which is the inward movement of the eyes towards the nose to focus on a common fixation point. It should be able to track and compute for the distance of a specific target object. It is composed of a computer which controls the whole system, two cameras that act as the 'eyes' and other control mechanisms for the movement of the eyes. The movement of both cameras, based on the principle of convergence, is the same with the movement of normal human eyes that is focused on an initially specified object. In this study, the visual system focuses on the tracking of a ball with a single distinct color and stationary background.