Mohd Romlay, Muhammad Rabani and Mohd Ibrahim, Azhar and Toha, Siti Fauziah and Rashid, Muhammad Mahbubur and Ahmad, Muhammad Syahmi (2022) Hardware development of autonomous mobile robot based on actuating lidar. ASEAN Journal of Engineering and Technology, 2 (3). pp. 1-10. E-ISSN 2600-9056
|
PDF
Download (510kB) | Preview |
Abstract
Object detection using a LiDAR sensor provides high accuracy of depth estimation and distance measurement. It is reliable and would not be affected by light intensity. However, high-end LiDAR sensors are high in cost and require high computational costs. In some applications such as navigation for blind people, sparse LiDAR point cloud are more applicable as they can be quickly generated and processed. As opposed to a point cloud generated from high-end LiDAR sensors where many algorithms have been developed for object detection, sparse LiDAR point clouds still possess large room for improvement. In this research, we present the construction of an autonomous mobile robot based on a single actuating LiDAR sensor, with human subjects as the main element to be detected. From here, the extracted values are implied on k-NN, Decision Tree and CNN training algorithm. The final result shows promising potential with 91% prediction when implemented on the Decision Tree algorithm based on our proposed system of a single actuating LiDAR sensor.
| Item Type: | Article (Journal) |
|---|---|
| Uncontrolled Keywords: | Obstacle avoidance, object recognition, autonomous mobile robot, decision tree. |
| Subjects: | T Technology > T Technology (General) > T173.2 Technological change |
| Kulliyyahs/Centres/Divisions/Institutes (Can select more than one option. Press CONTROL button): | Kulliyyah of Engineering Kulliyyah of Engineering > Department of Mechatronics Engineering |
| Depositing User: | Dr. Siti Fauziah Toha |
| Date Deposited: | 20 Jan 2023 12:03 |
| Last Modified: | 20 Jan 2023 12:04 |
| URI: | http://irep.iium.edu.my/id/eprint/103327 |
Actions (login required)
 |
View Item |
Download Statistics
Download Statistics