Design and Visio Control for Navigation and Obstacles Detection Based on Color and Texture Attributes of Two-Wheeled Mobile Robot
Sabri M. Ben Mansour1, Jawhar Ghommam2, Saber M. Naceur3
1Sabri M. Ben Mansour, LTSIRS, Remote Sensing and Spatially Referenced Information Systems Laboratory, National Engineering School of Tunis, Tunisia.
2Jawhar Ghommam, CEM-Laboratory, National Engineering School of Sfax, sfax, Tunisia. Saber
3M. Naceur, LTSIRS, Remote Sensing and Spatially Referenced Information Systems Laboratory, National Engineering School of Tunis, Tunisia.
Manuscript received on 20 May 2016 | Revised Manuscript received on 30 May 2016 | Manuscript published on 30 May 2016 | PP: 16-25 | Volume-5 Issue-2, May 2016 | Retrieval Number: B1572055216©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: This paper addresses Visio and path following control problem of a nonholonomic Two-Wheeled Inverted Pendulum Mobile Robot. We propose control architecture based on two control layers. A speed inner loop control scheme is first designed based on state feedback technique to ensure stability of the inverted structure of the robot. A second outer loop control scheme is proposed to help the robot navigate along a desired path formed by a set of way points. It is designed inspiring the model predictive control technique. The elements of the predictive control, which are the cost function, controls and constraints, must be defined and specified: the use of different trajectories group in the control can adapt the behavior of the robot to different displacement phases. The obstacle detection architecture based on the attributes of color and texture has been developed to be implemented on an Raspberry PI and is designed as a generic high-speed image processing device. The optimization criteria are based on a maximization of performance in terms of image processing per second and a minimization of consumed resources. Our obstacles detection algorithm consists of three main steps: the color transformation, the calculation of the color and texture attributes and their classification.
Keywords: Mobile robot, navigation, stability, Predictive control, Obstacles detections, Image processing
Scope of the Article: Signal and Image Processing