摘要
利用Android软硬件平台,设计并实现了一种融合颜色和形状特征的实时限速交通标志检测和识别系统。为了使检测和识别达到实时精确的标准,参考多种基于颜色和形状的检测算法以及分类算法,在检测阶段,选择在RGBN颜色空间进行分割后再结合本文提出的离心度的几何不变量进行形状筛选实现限速标志定位,在识别阶段,采用改进的动态阂值多模板匹配算法实现限速标志分类。实验表明,该检测和识别算法适用于移动平台,并且速度快,精度高。
Based on Android hardware and software platform,the article proposed a real-time speed limit traffic sign detection and recognition system which fused the color and shape features.In order to achieve high speed and accuracy,a variety of detection and classification algorithms based on color and shape features were referenced.In the detection phase,the RGBN color space was selected for fast segmentation,and the Contour Points' Distribution algorithm was proposed for fast shape screening to locate the traffic sign,in the recognition phase,the Dynamic Threshold Multi-templates Matching algorithm was proposed for fast recognition.The result turned out that the algorithms were suitable for android platform with high speed and accuracy.
出处
《微型电脑应用》
2016年第4期1-4,共4页
Microcomputer Applications
基金
国家自然科学基金重大研究计划集成项目(91220301)
关键词
交通标志
目标识别
颜色分割
模板匹配
Traffic Sign
Target Recognition
Color Segmentation
Template Match