Road Sign Alert and Drivers Drowsiness Alert using Convolutions Neural Network

  • M.S.Antony Vigil, Manoj Raj S R, Shankara Prasad S, Venkata Jagadeesh Reddy

Abstract

One of the most common causes of fatal vehicle accidents is drowsiness and misreading of road signs. As a result, there is always potential for development in the field of road sign identification and recognition approaches, and it has recently become one of the most prominent machine learning and computer vision projects. We present a model in this research that can recognise various sorts of traffic signals from photos. The dataset for this project is a widely used road sign. As a result, unlike many other projects, it is not restricted to a certain number of geographical areas. This venture might be utilized universally in light of the fact that the quantity of signs utilized for order is 43, which are utilized everywhere. We utilized two neural organizations, one for location and the other for acknowledgment; one distinguishes the sign and the other characterizes it. First classifier network was trained using 30k photos, containing 20k positive and 10k negative images, and the second classifier network was trained with 34k images. The photos are first analyzed to identify the region of interest, and then they are sent into the second network. This model could be utilized in assisted driving to keep drivers awake while also allowing the owners to see all the signs going on. We were able to achieve 98% percent accuracy by striking the correct balance between time and accuracy.

Published
2021-12-02
How to Cite
M.S.Antony Vigil, Manoj Raj S R, Shankara Prasad S, Venkata Jagadeesh Reddy. (2021). Road Sign Alert and Drivers Drowsiness Alert using Convolutions Neural Network. Design Engineering, 1602- 1609. Retrieved from http://www.thedesignengineering.com/index.php/DE/article/view/7114
Section
Articles