https://www.mdu.se/

mdu.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Real time image recognition of facial features for detecting a true breath
Mälardalen University, School of Innovation, Design and Engineering.
Mälardalen University, School of Innovation, Design and Engineering.
2015 (English)Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesis
Abstract [en]

The effects of driving under the influence of alcohol can be severe. In an effort to reduce the number of drunk drivers, alcolocks have been deployed. However, if a driver tries to cheat the system and succeeds, the effects may cause severe damage to both humans and property. One way of beating the system is by fooling the breathalyzer that the breath produced doesn’t contain alcohol. This is usually done using a device which either filters a breath, or by having a device that blows clean air into the alcolock. In this paper the problem of if it’s possible to determine whether a true breath is made by using facial and feature detection, is answered. A true breath is an unobstructed, real breath from a person. True breath identification can be used in order to, for example, determine if a mouthpiece-less alcolock device is being used correctly. To solve this problem facial and feature detection was used. In order to find the best methods for facial and feature detection, research result in this area was used. The method used to solve this problem works by first tracking the face. The mouth is then searched for in the lower part of the face region. After that the mouth is found, the lips are to be extracted from the mouth area using various methods. The lip tracking is then used to determine the mouth state, it does this by calculating the distance between the different points produced by the lip tracking. The mouth state, mouth and face tracking are then used during the identification process to determine if a real person was present and if he/she made a true breath. The test fails if the mouth is closed during the time of breath or if there is no mouth or face present. Attempts to manipulate, result in a failed alcohol test. In the majority of the test cases the tracking and mouth state scored acceptable or better results. The performance of the system concluded that it’s possible to complete the analysis in real time. The result of this thesis is a true breath identification system. It still has some issues in cases where the illumination and lighting is uneven or bad. Facial hair may also interfere with the tracking in some cases. Although there are still some areas that needs improving, this work was more of a proof of concept. This thesis shows that it is possible, under certain circumstances, to determine a true breath using facial and feature detection.

Place, publisher, year, edition, pages
2015.
National Category
Computer Sciences
Identifiers
URN: urn:nbn:se:mdh:diva-28358OAI: oai:DiVA.org:mdh-28358DiVA, id: diva2:822997
Subject / course
Computer Science
Supervisors
Examiners
Available from: 2015-06-30 Created: 2015-06-17 Last updated: 2018-01-11Bibliographically approved

Open Access in DiVA

bilaga(1773 kB)3446 downloads
File information
File name ATTACHMENT01.pdfFile size 1773 kBChecksum SHA-512
f0d9b9076387437b63ada0c62bf08ee4a6b47de0db2f88e0d46189610020e600f058f0041c2051d8fdff0feb2e4b36072b2644c42e8f2c30126569628dd28d49
Type attachmentMimetype application/pdf

By organisation
School of Innovation, Design and Engineering
Computer Sciences

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 358 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf