Real-time traffic sign recognition based on shape and color classification

摘要

In this work, a system is constructed to detect and classify traffic signs, aiming to build an early warning mechanism and assists the driver in recognizing basic and important signs. Recognition of all signs is beyond the scope of this work. The operation of this study is limited to traffic signs with specific shape and color under certain environmental circumstances. In this paper, detection of red and circular or triangular signs are targeted by collecting data using a camera, mounted in front of a vehicle. Traffic signs within the collected data are extracted from each frame through several detection phases and classified using Support Vector Machine algorithm which uses Histogram of Oriented Gradients as a feature.