Automated traffic sign recognition system using computer vision

Abstract

There are many existing companies who are developing cars on the autonomous driving technology. With the help of GPS and internet connectivity they create a dynamic map which helps the cars to navigate. This technology is still new and undergoing rigorous changes. There are many shortcomings to this existing technology. They are capable of navigating through those areas which are accounted for and surveyed but when the car enters in any unchartered terrain or there is any internet connectivity issues, the updation in the map is not possible, which leaves the car to navigate on its own. This can cause many troubles like you can get late or maybe lost. So to overcome these problems we need such an intelligent system with the help of camera feeds can monitor and identify the traffic signals dynamically. Traffic sign recognition is based on Advanced Driving Assistance System (ADAS) which is used by vehicles to recognise various traffic signs ahead. The system takes continuous video input from the dashboard camera or the camera mounted on the bonnet of the car. The underlying algorithm extracts the features of the input image and matches them with an existing library of traffic sign. The output is fed to the driving assistance system and it in turn drives the car accordingly. This intelligent system uses computer vision. This device will take camera feeds and upgrade the ADA system instantaneously. The algorithm has been implemented using Python language.