Traffic Sign Detection and Tracking Based on Shape and Color Information

概要

論文の詳細を見る
Traffic signs are the main guide of road transportation. To lighten the drivers' burden, vision based traffic sign recognition is developed as one important function in driver assistant system. In order to accurately locate traffic signs in continuous frames, traffic sign detection and tracking method is proposed in this paper. The shape and color information of traffic sign is used to detect traffic signs in single frame. To maintain detection result about traffic sign in continuous frames, Kalman filter is adopted in tracking step. The experiment results demonstrate that the proposed method can effectively detect traffic signs and track them.
社団法人電子情報通信学会の論文
2011-03-04