Traffic Symbol Recognition in Automotive Systems

Abstract

Road traffic sign boards are designed with purpose and it is meant to be followed by everyone driving on the road. As it can be useful in road traffic management and can help the driver to drive vehicle with certain discipline. But usually these signs get neglected from vision of the driver which leads to road traffic as well as road accidents. Hence to provide complete assistance to the driver for capturing and identifying various traffic symbols, a system is designed for traffic symbol recognition in automotive systems using Raspberry Pi. The Digital image processing plays important role in traffic symbol recognition system. To capture traffic signs, camera is being used which will be installed on the vehicle. Segmentation of traffic sign from captured image is done using canny edge detection method and also by examining the color using the color space. Then Principal Component Analysis (PCA) is used for feature extraction from segmented image. This algorithm gives precise results compared to other methods. Then feature matching is done using Euclidian distance. As a result, stipulated actions will be taken place automatically according to traffic symbol such as speed limit, horn control, etc. and also driver will get the information about that particular traffic sign through the LCD.