Artificial intelligence for cloud-assisted object detection
The research focuses on the integration of artificial intelligence (AI) and cloud computing to develop a License Plate Recognition (LPR) model. The existing LPR model is a combination of the detection (YOLOv4-tiny) model and the recognition (ResNet-FC) model. The LPR model is then deployed from the...
Saved in:
Main Author: | |
---|---|
Format: | Final Year Project / Dissertation / Thesis |
Published: |
2023
|
Subjects: | |
Online Access: | http://eprints.utar.edu.my/5646/1/3E_1804018_FYP_report_%2D_WEN_JIE_CHAN.pdf http://eprints.utar.edu.my/5646/ |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
id |
my-utar-eprints.5646 |
---|---|
record_format |
eprints |
spelling |
my-utar-eprints.56462023-07-06T12:06:06Z Artificial intelligence for cloud-assisted object detection Chan, Wen Jie TK Electrical engineering. Electronics Nuclear engineering The research focuses on the integration of artificial intelligence (AI) and cloud computing to develop a License Plate Recognition (LPR) model. The existing LPR model is a combination of the detection (YOLOv4-tiny) model and the recognition (ResNet-FC) model. The LPR model is then deployed from the local environment to the Intel® Developer Cloud for the Edge for further improvement (e.g. addition to the feature of selection of the inferencing engine provided by Intel). After deployment, this LPR model can be accessed from anywhere, anytime as long as internet connectivity is available. The purpose of the LPR model is to detect license plates in videos uploaded by users, reducing the need for manual monitoring and recording of car plate numbers. Next, further improvement was made by replacing the existing detection model (YOLOv4-tiny) with the more advanced version by using the YOLOv7 series. Subsequently, the detection model (YOLOv7-tiny) with the highest Mean Average Precision (mAP)@.5 of 0.936, and mAP@.5:.95 of 0.720, will replace the YOLOv4-tiny. Among the commonly used Intel hardware, the inference engine with Intel® Core™ i7-1185G7E and Intel® Iris® Xe Graphics 530 GPU integrated with the CPU had the highest performance. This inference engine had a processing time of 29 s and a frame per second (FPS) of 5.7 when applying to the LPR system with the YOLOv4-tiny detection model. 2023 Final Year Project / Dissertation / Thesis NonPeerReviewed application/pdf http://eprints.utar.edu.my/5646/1/3E_1804018_FYP_report_%2D_WEN_JIE_CHAN.pdf Chan, Wen Jie (2023) Artificial intelligence for cloud-assisted object detection. Final Year Project, UTAR. http://eprints.utar.edu.my/5646/ |
institution |
Universiti Tunku Abdul Rahman |
building |
UTAR Library |
collection |
Institutional Repository |
continent |
Asia |
country |
Malaysia |
content_provider |
Universiti Tunku Abdul Rahman |
content_source |
UTAR Institutional Repository |
url_provider |
http://eprints.utar.edu.my |
topic |
TK Electrical engineering. Electronics Nuclear engineering |
spellingShingle |
TK Electrical engineering. Electronics Nuclear engineering Chan, Wen Jie Artificial intelligence for cloud-assisted object detection |
description |
The research focuses on the integration of artificial intelligence (AI) and cloud computing to develop a License Plate Recognition (LPR) model. The existing LPR model is a combination of the detection (YOLOv4-tiny) model and the recognition (ResNet-FC) model. The LPR model is then deployed from the local environment to the Intel® Developer Cloud for the Edge for further improvement (e.g. addition to the feature of selection of the inferencing engine provided by Intel). After deployment, this LPR model can be accessed from anywhere, anytime as long as internet connectivity is available. The purpose of the LPR model is to detect license plates in videos uploaded by users, reducing the need for manual monitoring and recording of car plate numbers. Next, further improvement was made by replacing the existing detection model (YOLOv4-tiny) with the more advanced version by using the YOLOv7 series. Subsequently, the detection model (YOLOv7-tiny) with the highest Mean Average Precision (mAP)@.5 of 0.936, and mAP@.5:.95 of 0.720, will replace the YOLOv4-tiny. Among the commonly used Intel hardware, the inference engine with Intel® Core™ i7-1185G7E and Intel® Iris® Xe Graphics 530 GPU integrated with the CPU had the highest performance. This inference engine had a processing time of 29 s and a frame per second (FPS) of 5.7 when applying to the LPR system with the YOLOv4-tiny detection model. |
format |
Final Year Project / Dissertation / Thesis |
author |
Chan, Wen Jie |
author_facet |
Chan, Wen Jie |
author_sort |
Chan, Wen Jie |
title |
Artificial intelligence for cloud-assisted object detection |
title_short |
Artificial intelligence for cloud-assisted object detection |
title_full |
Artificial intelligence for cloud-assisted object detection |
title_fullStr |
Artificial intelligence for cloud-assisted object detection |
title_full_unstemmed |
Artificial intelligence for cloud-assisted object detection |
title_sort |
artificial intelligence for cloud-assisted object detection |
publishDate |
2023 |
url |
http://eprints.utar.edu.my/5646/1/3E_1804018_FYP_report_%2D_WEN_JIE_CHAN.pdf http://eprints.utar.edu.my/5646/ |
_version_ |
1772816670436360192 |
score |
13.18916 |