Please use this identifier to cite or link to this item:
Title: Object detection using artificial intelligence
Authors: Wang, Tian
Keywords: Engineering::Electrical and electronic engineering::Electronic systems::Signal processing
Issue Date: 2022
Publisher: Nanyang Technological University
Source: Wang, T. (2022). Object detection using artificial intelligence. Master's thesis, Nanyang Technological University, Singapore.
Abstract: Computer vision technology is changing the way people live. One important issue of computer vision is object detection which is the basis for high-level semantic information analysis of images. The objective of the object detection is to detect all items of the predefined classes and provide its localization by bounding boxes. It is a supervised learning problem. Object detection has many applications like face detection, vehicle detection, people counting, security and surveillance and so on. Integrating object detection technology into factory management has many benefits. It could help to monitor safety protection, improve production efficiency, control product quality and so on. Therefore, in my dissertation, I did detailed literature review of state of the art object detectors and did a comparison of the common methods. I also chose YOLOv5 as the candidate methods and did more evaluation of its models. At the same time, we made our own dataset including 3069 images and most of which were collected from the factory environment. After that, I trained YOLOv5 models on custom dataset on Google Colab and got excellent result. I also did some visualization and analysis of the the result and proposed some directions to improve the models in the future.
Schools: School of Electrical and Electronic Engineering 
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:EEE Theses

Files in This Item:
File Description SizeFormat 
  Restricted Access
2 MBAdobe PDFView/Open

Page view(s)

Updated on Feb 25, 2024


Updated on Feb 25, 2024

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.