Abstract
The purpose of this article is to present a theoretical and practical implementation of a fuzzy algorithm methodology to control a mobile manipulator path planning using a real-time vision system tracking. To meet high performance response and robust stability of the platform navigation, a fuzzy logic controller is designed with realistic constrains. OpenCV library is used to implement Background Modeling technique to track in real time a color object and to extract its (X, Z) coordinates, then an ultrasonic sensor is coupled with the camera to calculate the depth "Y" of the tracked object position. The inverse kinematics is used to control an arm robot to achieve a grasping task of the tracked object. The robot uses the vision system and the ultrasonic sensor to approximate the position of object compared to the cart as well as the position of the arm end effector to the target. The proposed technique shows through simulations and hardware implementation the high efficiency of the algorithm implemented. The fuzzy controller technique presents a good stability and robustness behavior results. The obtained results conclude that the combination between a 2D vision system and an ultrasonic sensor applied to a rigorous fuzzy logic algorithm can perform good results similar to a tracking technique based on a 3D camera.