Vision-Based Traffic Hand Sign Recognition for Driver Assistance

Abstract

Authorized traffic control has the highest priority in the case of high-traffic situations. In such situations, it is very important to have a smooth flow of traffic which is the job of ATC. ATCs make use of hand gestures for controlling traffic. It is very important for drivers to understand these gestures in order to follow their instructions otherwise it might lead to accidents, traffic jams, etc. However, most drivers nowadays are unaware of these signals and in the age of autonomous vehicles, it has become of utmost importance that the vehicles have a high level of understanding of these gestures to better assist the drivers. In this study, various approaches to this problem are tested. A detailed comparison of the proposed methods with previous works on this topic is done and the room for further improvement of the performance is discussed.