North-American speed limit sign detection and recognition for smart cars

Traffic sign detection and recognition system is becoming an essential component of smart cars. Speed-Limit Sign (SLS) is one of the most important traffic signs, since it is used to regulate the speed of vehicles in downtown and highways. The recognition of SLS by drivers is mandatory. In this paper, we investigate SLS detection and recognition system. We focus on North-American speed limit signs, including Canadian and U.S. signs. A modified version of Histogram of Oriented Gradients (HOG) is used to detect and recognize SLS through a set of two-level SVM-based classifiers. Moreover, we build our online database called North-American Speed Limit Signs (NASLS) which includes four SLS categories; white, yellow, black and orange signs. We show through an extensive set of experiments that our system achieves an accuracy of more than 94% of SLS recognition.

[1]  A. Herbin,et al.  Robust on-vehicle real-time visual detection of American and European speed limit signs, with a modular Traffic Signs Recognition system , 2007, 2007 IEEE Intelligent Vehicles Symposium.

[2]  K. Mohammadi,et al.  Speed limit traffic sign detection and recognition , 2004, IEEE Conference on Cybernetics and Intelligent Systems, 2004..

[3]  John D. Owens,et al.  Feature-based speed limit sign detection using a graphics processing unit , 2011, 2011 IEEE Intelligent Vehicles Symposium (IV).

[4]  法律.  Manual on Uniform Traffic Control Devices , 2010 .

[5]  Charles Wang,et al.  An efficient real time rectangle speed limit sign recognition system , 2010, 2010 IEEE Intelligent Vehicles Symposium.

[6]  John D. Owens,et al.  A Template-Based Approach for Real-Time Speed-Limit-Sign Recognition on an Embedded System Using GPU Computing , 2010, DAGM-Symposium.

[7]  C. Bahlmann,et al.  Real-time recognition of U.S. speed signs , 2008, 2008 IEEE Intelligent Vehicles Symposium.

[8]  Robert C. Bolles,et al.  Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography , 1981, CACM.

[9]  Wei Liu,et al.  U.S. speed limit sign detection and recognition from image sequences , 2012, 2012 12th International Conference on Control Automation Robotics & Vision (ICARCV).

[10]  Jim Tørresen,et al.  Recognizing Speed Limit Sign Numbers by Evolvable Hardware , 2004, PPSN.

[11]  Yonatan Wexler,et al.  Detecting text in natural scenes with stroke width transform , 2010, 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[12]  M.L. Eichner,et al.  Integrated speed limit detection and recognition from real-time video , 2008, 2008 IEEE Intelligent Vehicles Symposium.

[13]  J. Torresen,et al.  Efficient recognition of speed limit signs , 2004, Proceedings. The 7th International IEEE Conference on Intelligent Transportation Systems (IEEE Cat. No.04TH8749).

[14]  Fatin Zaklouta,et al.  Real-Time Traffic-Sign Recognition Using Tree Classifiers , 2012, IEEE Transactions on Intelligent Transportation Systems.

[15]  Bill Triggs,et al.  Histograms of oriented gradients for human detection , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[16]  Alain Rakotomamonjy,et al.  Object Categorization Using Kernels Combining Graphs and Histograms of Gradients , 2006, ICIAR.

[17]  B. Bradai,et al.  Improving pan-European speed-limit signs recognition with a new “global number segmentation” before digit recognition , 2008, 2008 IEEE Intelligent Vehicles Symposium.

[18]  A. Zelinsky,et al.  Real-time radial symmetry for speed sign detection , 2004, IEEE Intelligent Vehicles Symposium, 2004.