摘要
针对现有道路交通标志图像分割算法中实时性和鲁棒性较差的问题,提出一种基于免疫算法和RGB颜色空间的交通标志图像分割新方法。该方法先用RGB颜色空间将交通标志从实景图中定位出来,利用免疫算法快速搜索的机理,对已定位的图像进行阈值T寻优,并用该阈值T分割图像。实验结果表明,该算法与基于遗传算法的图象分割算法相比较,具有较好的鲁棒性和实时性,且适用于不同光照条件下的交通标志图像分割。
Some existing traffic signs graphic segmentation algorithms are weak in perform real-time and robust. According to those problems, this paper presents a new algorithm of the traffic signs segmentation based on immune algorithm RGB color space. The method uses RGB color space to locate the image of traffic signs form the traffic signs imaging figure, and use immune algorithm to search for the optimization threshold T to split the located image. The experimental results show that, compared with the algorithm of genetic-based traffic signs segmentation, the algorithm presents perform more perfectly in robust and real- time ,and applies to traffic signs segmentation in varies light conditions.
出处
《信息技术》
2013年第7期54-57,62,共5页
Information Technology
基金
广西科技大学(筹)科学基金资助项目(校科自1261104)
广西自然科学基金项目(2010GXNSFA013126)
关键词
交通标志
免疫算法
颜色空间
图像分割
最优阈值
traffic signs
immune algorithm
color space
image segmentation
optimal threshold
