Traffic light detection for colorblind individuals
Jamal Al-Nabulsi, Abdelwadood Mesleh, Adnan Yunis · 2017
This paper proposed an algorithm that detects traffic light colors for colorblind individuals, the proposed algorithm employs image processing techniques associated in image processing toolbox in LabVIEW to help colorblind individuals in identifying the colors of traffic lights. It uses a fixed mobile camera to capture traffic light images taken in different roads and streets in Jordan and Kuwait. It detects traffic lights by comparing the candidate traffic light with some in-house collected traffic light templates, comparison is based on correlation. The templates represent 22 different shapes of traffic lights in Jordan and Kuwait. Finally, the algorithm extracts the green and the red planes and recognizes their colors. Experimental results reveal the accuracy of proposed algorithm in identifying the colors of traffic lights in different cases and circumstances. Hence, our proposed algorithm is helpful for colorblind drivers.