Paper Details

This project involves comparative analysis of detection of traffic light signals with the use of image processing..This paper focuses on a product of usage of monocular video camera for the detection of traffic lights under differentconditions i.e. adverse weather and low illumination, haze, fog etc. This paper provides a perception algorithm that works atreal time with high apprehension rate while using closeout camera for traffic light signal detection. By the use of thisalgorithm credibility and immunity for UGV and drivers will step-up. The working of traffic light detection system is dividedinto in three parts i.e. a monocular or color camera, then acquisition of image by that camera and a PC is required toprocess the captured images. The RGB color space based algorithm helps to detect and extract the red, green and yellowobjects in the images and then illuminating the disturbing colors present in the locale. The final objective is to verify thefeatures and object identity and based on these the different types of traffic signals are examined. . The ideology isqualitatively computed in various conditions, including driving in the heavy rain, at night and in city roads with dense traffic,as well as their synergy. Quantitative computation on a publicly available manually commented database, scoring highapprehension rate is also implemented.

Aradhana Verma, Dr.C.S.Yadav, “Traffic Light Signal Detection: A Study”, International Journal of Advance Engineering and Research Development (IJAERD), Vol. 4, Issue 8, pp. 59-63, August 2017.