Analys av NIR hårdvaruuppsättning samt metod för fotografering av individer i fordon
Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesis
Purpose – To study the possibility of using a NIR hardware solution to photograph individuals in a private vehicle as well as an analysis of its images.
Method – A study of existing theories around the NIR-method’s performance in selected conditions and individual tests were performed to examine if the literature statements were valid for this study. Two empirical tests have been carried out, the first was carried out at Kapsch test track and the other outside the test track in a single stationary test. An interview which formed the basis for the assessment on the quality of empirical data with a focus on computer-based detection of the number of individuals in the vehicle.
Findings – The results have demonstrated the potential of the NIR method’s performance in a fully automated detection system for the number of individuals inside the vehicle. Empirical data indicates that the method can depict individuals inside vehicles of sufficiently high quality, but it is greatly affected by reflections, weather and light conditions.
Implications – Result supports the assumption that the NIR method using an external light source can be used to image the interior through a varying number of weather and lightning conditions. The study originated until the results suggested that a NIR-based hardware setup can create images with high enough quality for the human eye to be able to detect the number of individuals inside the vehicle.If the overall performance into account, it suggests that the main problem with the use of the hardware set is to maintain the quality of the whole sample and that the crucial variables for the method’s performance is the influence of light and reflection conditions.
Limitations – The major limitations have been that we limited ourselves to a subjective analysis of the selection and assessment on the image features for computer-based detection of the number of individuals in the vehicle. We were limited to two tests, one in tough conditions where only the driver was in the vehicle and the second stationary test, where the focus was on the number of people in vehicles and light sources impact on the result.
Place, publisher, year, edition, pages
2016. , 43 p.
NIR, IR, HOV, bildanalys, trafikövervakning
IdentifiersURN: urn:nbn:se:hj:diva-32265ISRN: JU-JTH-DTA-1-20160030OAI: oai:DiVA.org:hj-32265DiVA: diva2:1046305
Subject / course
JTH, Computer Engineering