AI-based drone vision and control

Artificial intelligence has been the backbone of many industry leading application. It makes electronics equipped with it smarter allowing the users to tap on its ability to process information quick. What if it was used to aid vision equipped device to be able to assist human user improve their eve...

Full description

Saved in:
Bibliographic Details
Main Author: Pua, Lennard Wei Jian
Other Authors: Wen Bihan
Format: Final Year Project
Language:English
Published: Nanyang Technological University 2024
Subjects:
Online Access:https://hdl.handle.net/10356/177235
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:Artificial intelligence has been the backbone of many industry leading application. It makes electronics equipped with it smarter allowing the users to tap on its ability to process information quick. What if it was used to aid vision equipped device to be able to assist human user improve their everyday lives as well as their safety? This paper aims to explore the possibility of AI vision equipped drone helping lifeguards to patrol the pools and the sea to help identify swimmers at risk of drowning. It will help to evaluate the model of choice, YOLOv8’s performance as a real time object detection model on a drone and it also serves as a documentation of the considerations of this project as well as its difficulties faced while developing the tracking algorithm as well as the controller before concluding on the possibility of deploying the model in real applications.