Vehicle Brake Light Detection Using Hybrid Color Model

Mohd Fidzrin, Mohd Hamizaz (2014) Vehicle Brake Light Detection Using Hybrid Color Model. Project Report. UTeM, Melaka, Malaysia. (Submitted)

Text (24 Pages) Vehicle Brake Light Detection Using Hybrid Color Model 24 Pages.pdf - Submitted Version Download (261kB)

Abstract

Vehicle brake light detection is an important safety feature that should have in every car in Malaysia. By refer to the statistic of accidents in Malaysia by the Ministry of Transportation Malaysia in 2009 to 2012 is increasing, the total number of road accidents in Malaysia are 341 252, 363 319, 373 071, 397 330, 414 421, 449 040 and 462 423 respectively. Therefore, we propose a project to detect vehicle brake light to reduce number of accident in Malaysia. This project can be applied in intelligent vehicles and able to give an early warning to the driver in the future if someone is breaking ahead. On top of it, this system also can provide safe feeling and confident while driving to the driver. This project is focusing in Melaka region only and four various weather will be considered in image capturing. MATLAB will be used to develop an algorithm to detect vehicle brake light. Furthermore, in order to detect vehicle brake light, a hybrid color model will be applied which consist of combination of red color from RGB color model and method from HSI color model. At the end of this project, we are able to detect the vehicle brake light in 3.32 seconds delay and 83.31% of detection.

Actions (login required)

View Item