Author/Authors :
Khusheef, Ahmed S. Edith Cowan University - School of Engineering, Australia , Kothapalli, Ganesh Edith Cowan University - School of Engineering, Australia , Tolouei-Rad, Majid Edith Cowan University - School of Engineering, Australia
Abstract :
This study proposes a vision-based methodology for searching and relocation of target objectsusing a hexapod mobile robot. Such robot needs enhanced systems for navigation and visionbasedobject recognition. The navigation system is important for generating a path that covers the entire environment and for locating the position of the robot within that environment. Vision system is also important for undertaking the exploration task; it involves using a vision sensor and employing an optimal object recognition technique. This paper explains the development of such systems and presents a strategy for constructing the sensory platform to enable the robot to navigate and explore its environment. It also explains the challenges that arise from the viewpoint of vision sensor to the objects and suggests some relevant solutions. The experimental results demonstrate that it is possible to implement a navigation system within a minimum number of sensors if they are properly positioned in the robot’s body. The Experiments proved that the methodology used and that the codes developed made the robot capable of performing its task of finding, approaching and relocating objects as planned.
