Real-time traffic sign recognition based on Zynq FPGA and ARM SoCs

In this paper, an FPGA-based traffic sign recognition system is introduced for driver assistance applications. The system incorporates two major operations, traffic sign detection and recognition. The algorithms presented include hue detection for potential sign detection, morphological filters for noise reduction, labeling and Hausdorff distance calculation for template recognition. A new hardware platform is presented that combines a Zynq-7000 FPGA processing system and custom IP peripherals together. A frame-work for embedded system development on ARM CPU cores and FPGA fabric is introduced. The proposed hardware platform achieves up to 8 times speed-up compared to the existing FPGA based solutions.

[1]  C. Beleznai,et al.  Road Sign Detection from Edge Orientation Histograms , 2007, 2007 IEEE Intelligent Vehicles Symposium.

[2]  Luc Van Gool,et al.  SURF: Speeded Up Robust Features , 2006, ECCV.

[3]  Sebastian Thrun,et al.  Towards fully autonomous driving: Systems and algorithms , 2011, 2011 IEEE Intelligent Vehicles Symposium (IV).

[4]  T. Sledevic,et al.  SURF algorithm implementation on FPGA , 2012, 2012 13th Biennial Baltic Electronics Conference.

[5]  L. Boussaid,et al.  An embedded system for real-time traffic sign recognizing , 2008, 2008 3rd International Design and Test Workshop.

[6]  Wolfgang Rosenstiel,et al.  Design of an automotive traffic sign recognition system targeting a multi-core SoC implementation , 2010, 2010 Design, Automation & Test in Europe Conference & Exhibition (DATE 2010).

[7]  Thomas B. Moeslund,et al.  Vision-Based Traffic Sign Detection and Analysis for Intelligent Driver Assistance Systems: Perspectives and Survey , 2012, IEEE Transactions on Intelligent Transportation Systems.

[8]  Luc Van Gool,et al.  Speeded-Up Robust Features (SURF) , 2008, Comput. Vis. Image Underst..

[9]  Erdal Oruklu,et al.  FPGA-Based Traffic Sign Recognition for Advanced Driver Assistance Systems , 2013 .