We present a new modular traffic signs recognition system, successfully applied to both American and European speed limit signs. Our sign detection step is based only on shape-detection (rectangles or circles). This enables it to work on grayscale images, contrary to most European competitors, which eases robustness to illumination conditions (notably night operation). Speed sign candidates are classified (or rejected) by segmenting potential digits inside them (which is rather original and has several advantages), and then applying a neural digit recognition. The global detection rate is ~90% for both (standard) U.S. and E.U. speed signs, with a misclassification rate <1%, and no validated false alarm in >150 minutes of video. The system processes in real-time ~20 frames/s on a standard high-end laptop.
updated: Wed Oct 07 2009 15:43:01 GMT+0000 (UTC)
published: Wed Oct 07 2009 15:43:01 GMT+0000 (UTC)