一种快速的交通标志检测算法

A Fast Traffic Sign Detection Algorithm

交通标志识别包括交通标志检测和交通标志分类两个步骤,而决定交通标志识别实时性的关键在于交通标志的检测这一步骤.如何快速检测可能出现的交通标志的区域是实时交通标志识别的关键.本文就目前交通标志检算法普遍存在实时性不足的问题,提出了基于颜色概率模型和BING的快速交通标志检测算法.实验结果表明,本文提出的算法能快速地从待检测图像中筛选出包含交通标志的尽量少的候选窗口,不需要处理传统的滑动窗口方法产生的数万或数十万窗口,进而减少整个交通标志检测的时间,达到实时检测交通标志的要求.

In general, traffic recognition consists of two steps： traffic sign detection and traffic sign classification. Traffic sign detection is the key step in a real-time traffic sign recognition system. Addressing the problem that most existing methods cannot be implemented in real time, this paper presents a fast traffic sign detection algorithm based on a color probability model and BING. The experimental results show that the proposed algorithm can select as few candidate windows containing traffic signs as possible from the these detected images quite fast, which avoids handling tens of thousands windows as done in traditional sliding windows methods, hence, the computational cost of the proposed algorithm is reduced so that the real-time detection requirement is achieved.