Car cabin surveillance using computer vision

Driver assistance systems (DAS) in cars has become more intelligent than ever and it has enabled autonomous driving in certain conditions. Consequently, there is a need to always monitor the drivers’ attention inside the car to make sure that they are attentive enough and ready to take over when the...

Full description

Saved in:
Bibliographic Details
Main Author: Soegeng, Andrew Ivan
Other Authors: Lap-Pui Chau
Format: Final Year Project
Language:English
Published: Nanyang Technological University 2022
Subjects:
Online Access:https://hdl.handle.net/10356/157277
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-157277
record_format dspace
spelling sg-ntu-dr.10356-1572772023-07-07T19:00:53Z Car cabin surveillance using computer vision Soegeng, Andrew Ivan Lap-Pui Chau School of Electrical and Electronic Engineering Continental Guo Heng asoegeng001@e.ntu.edu.sg, elpchau@ntu.edu.sg Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence Driver assistance systems (DAS) in cars has become more intelligent than ever and it has enabled autonomous driving in certain conditions. Consequently, there is a need to always monitor the drivers’ attention inside the car to make sure that they are attentive enough and ready to take over when the DAS asked. As such, this research proposes a new computer vision approach of where there would be a wide-angle camera placed in the center mirror monitoring the driver and passenger and it will be connected to a computing unit inside the car which utilizes deep learning-based AI model to detect the activity of the driver and passenger and determine whether it is safe to engage level 2 DAS. In this paper, a literature review about the different approaches to the problem will be given together with the different datasets that are available to support the model training. Moreover, this paper presents the findings and the results from 2 different approaches which are classic image classification-based approach, and a novel 2-stage classifier was proposed whereby the frames are passed to pose estimation and face mesh detection model that will output key points for both the face mesh and the human body which are then passed to an algorithm that will output the driver activity. Bachelor of Engineering (Electrical and Electronic Engineering) 2022-05-13T07:12:12Z 2022-05-13T07:12:12Z 2022 Final Year Project (FYP) Soegeng, A. I. (2022). Car cabin surveillance using computer vision. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/157277 https://hdl.handle.net/10356/157277 en B3039-211 application/pdf Nanyang Technological University
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
spellingShingle Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
Soegeng, Andrew Ivan
Car cabin surveillance using computer vision
description Driver assistance systems (DAS) in cars has become more intelligent than ever and it has enabled autonomous driving in certain conditions. Consequently, there is a need to always monitor the drivers’ attention inside the car to make sure that they are attentive enough and ready to take over when the DAS asked. As such, this research proposes a new computer vision approach of where there would be a wide-angle camera placed in the center mirror monitoring the driver and passenger and it will be connected to a computing unit inside the car which utilizes deep learning-based AI model to detect the activity of the driver and passenger and determine whether it is safe to engage level 2 DAS. In this paper, a literature review about the different approaches to the problem will be given together with the different datasets that are available to support the model training. Moreover, this paper presents the findings and the results from 2 different approaches which are classic image classification-based approach, and a novel 2-stage classifier was proposed whereby the frames are passed to pose estimation and face mesh detection model that will output key points for both the face mesh and the human body which are then passed to an algorithm that will output the driver activity.
author2 Lap-Pui Chau
author_facet Lap-Pui Chau
Soegeng, Andrew Ivan
format Final Year Project
author Soegeng, Andrew Ivan
author_sort Soegeng, Andrew Ivan
title Car cabin surveillance using computer vision
title_short Car cabin surveillance using computer vision
title_full Car cabin surveillance using computer vision
title_fullStr Car cabin surveillance using computer vision
title_full_unstemmed Car cabin surveillance using computer vision
title_sort car cabin surveillance using computer vision
publisher Nanyang Technological University
publishDate 2022
url https://hdl.handle.net/10356/157277
_version_ 1772828907426283520