基于实景图像的道路限速标志检测算法研究

Research on Traffic Speed Limit Signs Location Algorithm Based on Real Image

在基于道路图像的交通标志识别系统中,关键步骤之一就是对图像中的交通标志能够快速有效地检测.以分析限速标志为例,根据限速标志的特点在图像中寻找到其所在的位置并提取出相关区域,为图像的识别做好前期准备.为了提取出交通标志,采用了颜色预处理、形态学滤波、二值化后的连通域处理和圆的检测等方法.实验表明,该方法能有效地从实景图像中提取出含有交通标志的区域.

One of the key steps is that the traffic signs can be located quickly and effectively in the road traffic sign recognition system based on road image. Based on the analysis of the speed limit signs as example, according to the characteristics of speed limit signs in the image, the location of the signs are found and the relevant areas are extrated, as the preparatory to image recognition. In order to extract the traffic signs, color processing, morphology filter, connected component disposal in binary images and circle detection are used. Experiments show that these methods can extract effectively the region containing traffic are used from real image.