摘要
考虑到交通标志具有颜色与形状这两个特征,提出了一种具有Harris角点特征的随机抽样一致性(RANSAC)算法来对交通标志进行检测。通过对含交通标志图像进行颜色特征提取后利用Canny算子对边缘进行检测,再利用Harris角点检测算子来提取初步检测后图像的角点,最后通过RANSAC算法与检测模型进行匹配进行检测,得出结果。仿真实验表明:该算法在检测率上达到了84%,也基本满足了实时性的要求。
In consideration of two features of traffic signs,colour and shape,a random sampling consensus（ RANSAC） which posses Harris angular point is proposed to detect traffic signs. The first phase is to conduct detection through Canny operator after colour extraction of the image. The second phase includes extracting initial image angular point with the help of Harris corner detector. The result is achieved by the matching between RANSAC algorithm and detection model. The simulation experiment shows that the 84 % of detection rate,basically meet the demand of instantaneity.
作者
戈侠
于凤琴
陈莹
GE Xia YU Feng-qin CHEN Ying(School of Internet of Things Engineering, Jiangnan University, Wuxi 214122, China)
出处
《传感器与微系统》
CSCD
2017年第3期124-127,共4页
Transducer and Microsystem Technologies
基金
国家自然科学基金资助项目(61104213)
