辅助驾驶中的红绿灯识别及其FPGA实现  

Traffic light recognition in auxiliary driving and its implementation by FPGA

作　　者：翟社平[1] 李威[1] 马蒙雨 

机构地区：[1]西安邮电大学计算机学院,陕西西安710121

出　　处：《现代电子技术》2016年第6期73-76,共4页Modern Electronics Technique

基　　金：陕西省自然基金面上项目(2012JM8044);陕西省教育厅项目(12JK0733);西安邮电大学创新基金项目(114-602080034);3S杯全国大学生物联网技术与应用项目(B149)

摘　　要：为了降低路况的复杂性和交通信号灯的多变性给司机带来的影响,模拟一种辅助驾驶情景,提出基于FPGA的颜色识别系统的设计,作为辅助驾驶系统中的交通灯识别模块。在识别系统中,使用颜色特征HSV作为特征提取算法。为了增强对交通灯识别的准确率,先对图像进行腐蚀操作,去掉小的噪点;然后再对图像进行膨胀操作,恢复对非噪声区域削去的部分边缘。实验结果表明,基于FPGA的识别模块可快速准确地识别红、黄、绿3种颜色,识别率达到97%。To reduce the effect of the complexity of road situation and polytropism of traffic lights on drivers,the design scheme of a color recognition system based on FPGA is proposed by simulating an auxiliary driving scene,which is taken as traffic light recognition module in driver assistance system. In recognition system,color feature-HSV is used as feature extraction algorithm. In order to enhance the traffic light recognition accuracy,the image etching is carried out first to remove the small noise,and then the image expansion operation is conducted to restore some edge of non-noise area,which is chipped off before.The experiment results show that FPGA-based identification module can quickly and accurately identify the three colors of red,yellow and green. Its recognition rate can reach up to 97%.

关 键 词：FPGA HSV 图像的腐蚀 颜色识别系统 

分 类 号：TN911-34[电子电信—通信与信息系统]