Giving smart devices a body
Monocular vision camera has been increasingly employed in autonomous vehicle navigation. For example, keeping the optical flow divergence at a constant value is one of the most popular methods to ensure Micro Air Vehicles' smooth landing. However, methodologies for robust control of autonomous...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project |
| Language: | English |
| Published: |
Nanyang Technological University
2021
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/151035 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | Monocular vision camera has been increasingly employed in autonomous vehicle navigation. For example, keeping the optical flow divergence at a constant value is one of the most popular methods to ensure Micro Air Vehicles' smooth landing. However, methodologies for robust control of autonomous car navigations are still underdeveloped. This is particularly true for Automated Guided Vehicles (AGV), for which the surrounding environment can be much noisier and more dynamic than Unmanned Aerial Vehicle (UAV). In this project, a smartphone device is directly connected to an omnidirectional NEXUS ROBOT. The built-in monocular camera from the smartphone is used to capture the surrounding environment of the vehicle. The objective of this project is to study the optical flow control method for autonomous vehicle navigation. Firstly, a brief introduction and literature review of the optical flow control strategy is conducted. A detailed explanation of optical flow control algorithms and their implementations are further demonstrated, including Harris and Shi-Tomasi corner detection algorithms, Horn-Shunck method, Gunnar Farneback method and Lucas-Kanade method. In addition, a case study of distance estimation with an extended Kalman filter based on previous research is demonstrated to validate the effectiveness of optical flow control method. The computer simulation is conducted using MATLAB Simulink to perform the nonlinear system state estimation using the EKF method. Furthermore, Python OpenCV is used to implement the optical flow estimation methods with an example. Lastly, potential commercial applications and future research perspectives is proposed based on optical flow control for autonomous vehicle navigation. |
|---|