Authors : Hairol Nizam Mohd Shah, Marizan Sulaiman, Mohd Zamzuri AB Rashid, Zalina Kamis, Nursabillilah Mohd Ali, Mohd Shahrieel Mohd Aras, Faizil Wasbari and Zakarya Omar
Abstract: In this study, a 3D mapping of the surrounding object using stereo-vision technique is proposed. A stereo-vision system is constructed and a navigation algorithm is developed to extract 3D features of objects in the surrounding. Those 3D features are back projected onto a mesh grid to reconstruct the 3D module of that object. Building 3D maps and autonomous localization is a fundamental characteristic of an autonomous operating robot in unknown environment. In addition, rescue activities in hazardous places require robots with such capabilities. In the past, building 3D maps were based on the use of laser, sonar sensor or a combination of a single camera with either laser or sonar sensor. Stereo vision system is the technique used to construct the 3D modules. Two cameras are mounted horizontally on the top of the robot and are directed to scan the forward path of the robot. 3D maps containing the location of objects in the surrounding provide the robot with a proper guidance necessary for its navigation. Those maps will enhance the vision-based robot’s ability to navigate and localize themselves autonomously in the future.
Hairol Nizam Mohd Shah, Marizan Sulaiman, Mohd Zamzuri AB Rashid, Zalina Kamis, Nursabillilah Mohd Ali, Mohd Shahrieel Mohd Aras, Faizil Wasbari and Zakarya Omar, 2017. A 3D Mapping of the Surrounding Object using Stereo-Vision Technique. International Journal of Soft Computing, 12: 59-65.