Methods and Means of Car Driver Decision Support for Speed Limitation

Abstract

This article is about road signs determination. The urgency of this task is determined by the issues of road safety. The development of modern computer technology has allowed many car manufacturers to establish vision systems in production cars. Over the past few years, computer vision has gained immense popularity. One of the tasks of computer vision is image recognition. However, the main problems of such systems are low detection accuracy, as well as the inability of some systems to recognize Russian traffic signs. The description of the road signs recognition system RoadAR based on Android is presented, as the most budgetary and affordable solution of the problem. The RoadAR system was tested in clear, cloudy weather and at night. As a result it was concluded that most systems recognizes limiting speed road signs, but do not control the zone of the sign. In this regard, it is necessary to develop algorithms for recognizing road signs that cancels the sign “Limitation of the maximum permissible speed”. The paper presents a general algorithm for character recognition using a Gaussian filter, binarization, the search for geometric shapes, and comparison with a standard. In addition, algorithms for recognizing the sign “The Beginning of the Settlement”, based on the Canni algorithm, segmentation and character recognition on the sign, is presented. The information on the recognized signs will be further used in the driver’s decision support system. The driver decision support algorithm is based on traffic rules, recognized signs and vehicle speed data.