| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 743595 | Optics and Lasers in Engineering | 2014 | 11 Pages |
•This paper aims to propose a robust and precise way to obtain a digitized map of robot's surrounding with metrological quality by means of optical measurements.•Our vision system has a 98% confidence level on its coordinate's measurements.•Two principal advantages are assumed over other known robot navigators, smoother and significantly shorter trajectory.•The use of Levenberg–Marquardt method helped us to increase the metrological accuracy of measured 3D coordinates up to 5%.
In our current research, we are developing a practical autonomous mobile robot navigation system which is capable of performing obstacle avoiding task on an unknown environment. Therefore, in this paper, we propose a robot navigation system which works using a high accuracy localization scheme by dynamic triangulation. Our two main ideas are (1) integration of two principal systems, 3D laser scanning technical vision system (TVS) and mobile robot (MR) navigation system. (2) Novel MR navigation scheme, which allows benefiting from all advantages of precise triangulation localization of the obstacles, mostly over known camera oriented vision systems. For practical use, mobile robots are required to continue their tasks with safety and high accuracy on temporary occlusion condition. Presented in this work, prototype II of TVS is significantly improved over prototype I of our previous publications in the aspects of laser rays alignment, parasitic torque decrease and friction reduction of moving parts. The kinematic model of the MR used in this work is designed considering the optimal data acquisition from the TVS with the main goal of obtaining in real time, the necessary values for the kinematic model of the MR immediately during the calculation of obstacles based on the TVS data.
