| Author: | Feng, Yurong |
| Title: | A real-time object inspection system for unmanned aerial vehicles |
| Advisors: | Wen, Chih-yung (ME) |
| Degree: | M.Sc. |
| Year: | 2020 |
| Subject: | Drone aircraft -- Automatic control Embedded computer systems Hong Kong Polytechnic University -- Dissertations |
| Department: | Department of Mechanical Engineering |
| Pages: | 74 pages : color illustrations |
| Language: | English |
| Abstract: | Visual inspections act as a foundation for maintenance of public facilities to preserve the proper and stable operation of facilities. Current methods for inspection public facilities are costly, lengthy and labor intensive. Nowadays, unmanned aerial vehicles (UAV) equipped with computer vision techniques provide a potential as a helpful tool for facility inspection. This study proposes an autonomous UAV system that detects and inspects specific facility. The system detects objects using a pre-trained deep learning model. We collect and generate a medium scale of images dataset and use it to train a set of object detection model. The best performance model integrated with a vision-based control path planning algorithm was examined on a physical UAV platform. Our system was able to successfully detect and inspect an object autonomously during flight test which validates our path planning algorithm. Detection performance was achieved by the detection algorithm's 93% accuracy in messy environment and the inference time of 2 seconds on a companion computer. Visual-servo path planning algorithm was justified in terms of the estimated objects position and dynamic characteristic of UAV. |
| Rights: | All rights reserved |
| Access: | restricted access |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| 5235.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 3.29 MB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/10798