基于日间行车的灯语识别技术

摘要/Abstract

摘要： 汽车的车灯不仅具有照明作用,也是车辆行驶时与周围车辆交流信息的重要途径。在辅助驾驶中,准确理解周围车辆传递出的灯语信息,是制定车辆正确驾驶决策的前提。在日间行车时,由于行车环境多变,单通过特征匹配的方式检测车灯或车辆,进而识别灯语的方式,很难在道路实测中取得良好的效果。为此,针对日间行车情况,文中提出一种基于车辆检测的灯语识别方法。首先,文中使用Haar特征结合Adaboost级联分类器的车辆检测方式进行车辆检测,并在此基础上根据车灯在车尾的位置分布特征确定感兴趣区域;然后,在RGB色彩空间中提出一种颜色分割算法,其能够在感兴趣区域的基础上精确提取车灯位置并判断车灯的点亮状态,同时使用车灯点亮时的亮度特征排除颜色分割算法导致的误检;最后,使用高位刹车灯作为刹车灯灯语的识别条件,将历史频率信息作为转向灯灯语的识别条件,完成了日间行车时前车尾灯灯语的识别。以VS2010和opencv3.4.9作为算法的实现工具,将上汽提供的行车记录仪中的道路实测数据进作为测试数据进行实验。经测试,使用更新样本的训练方法得到的分类器识别准确率为93%,相对于传统Adaboost分类器,识别准确率提升了约2%,灯语识别算法的平均精度为93%,其总体平均耗时约53ms。实验结果表明,分类训练方法能够小幅度提升检测精度,而灯语识别算法能够较准确地识别出刹车灯和转向灯以及两种灯语同时存在的情况,且基本保证实时性。

关键词: 车辆检测, 灯语识别, 辅助驾驶, 机器视觉, 颜色分割

Abstract: The car lights not only have lighting functions,but also are important ways for vehicles to communicate with other vehicles while driving.In assisted driving,understanding the light message transmitted by surrounding vehicles accurately is a prerequisite for making correct driving decisions.During daytime driving,due to thechangeble environment,it is difficult to achieve good results in road measurement by matching the lights and then recognizing the lamp language.To this end,in view of the daytime driving situation,this paper proposed a method of light language recognition based on vehicle detection.In this paper,the Adaboost cascade classifier is trained to test the vehicle by using the training method of the updated sample.Based on this,the position distribution feature of the vehicle rear is used to determine the region of interest of the lights.In the RGB color space,a color segmentation algorithm is proposed,which can accurately extract the position of the lamp,judge the lighting state of the lamp on the basis of the region of interest,and eliminate the misdetection of color segmentation algorithm.This paper uses the brightness feature when the lamp is lit.The high-position brake light is used as the recognition condition of the brake light lamp language,and the historical frequency information is used as the recognition condition of the turn signal light,and the recognition of the front taillight light during daytime driving is completed.The experiment uses VS2010 and opencv3.4.9 as the algorithm implementation tool,and uses the actual measured data of the driving recorder provided by SAIC as the test data.After test,the accuracy of classifier recognition in experimental training is 93％.Compared with the traditional Adaboost classifier,the recognition accuracy is improved by about 2%,the average accuracy of the light recognition algorithm is 93％,and the average time of the algorithm is about 53ms.The test results show that the classification training method used in the experiment can improve the detection accuracy slightly,and the light recognition algorithm can accurately identify the brakes,the turn signals and the two kinds of lights simultaneously,and can basically guarantee the real-time performance.

Key words: Assisted driving, Color segmentation, Computer vision, Lamp language recognition, Vehicle detection

中图分类号:

TP751

李堃, 黎向锋. 基于日间行车的灯语识别技术[J]. 计算机科学, 2019, 46(11A): 277-282. https://doi.org/

LI Kun, LI Xiang-feng. Lamp Language Recognition Technology Based on Daytime Driving[J]. Computer Science, 2019, 46(11A): 277-282. https://doi.org/

参考文献

[1]赵阳,刘红娅.车辆灯语检测与识别技术研究与仿真[J].装备学院学报,2012,23(3):93-97.
[2]田强.车辆尾灯的检测与灯语识别[D].合肥:中国科学技术大学,2015.
[3]李仪.城市环境中交通对象检测与识别研究[D].长沙:中南大学,2013.
[4]FRÖHLICH B,ENZWEILER M,FRANKE U.Will this carchange the lane?－Turn signal recognition in the frequency domain[C]∥Intelligent Vehicles Symposium Proceedings,2014 IEEE.IEEE,2014:.37- 42.
[5]毛世榕,胡桂明,胡进.基于单目视觉的车辆检测算法研究[J].网络安全技术与应用,2010(12):34-37.
[6]杨建荣,曲仕茹.基于单目视觉的障碍物检测方法研究[J].计算机仿真,2009,26(2):278-281.
[7]张建明,张玲增,刘志强.一种结合多特征的前方车辆检测与跟踪方法[J].计算机工程与应用,2011,47(5):220-223.
[8]鲁威威,肖志涛,雷美琳.基于单目视觉的前方车辆检测与测距方法研究[J].电视技术,2011,35(1):125-128.
[9]付彦君,裘伟,戴斌,等.基于多特征融合的车辆实时检测[J].计算机仿真,2007(10):212-215.
[10]SUN Z,BEBIS G,MILLER R.On-road vehicle detection using evolutionary gabor filter optimization[J].IEEE Transations on Intelligent Transportation Systems,2005,6(2):125-137.
[11]罗瑞奇,钟忺,钟珞,等.一种改进Haar-like特征的车辆识别算法[J].武汉大学学报(理学版),2018,64(3):244-248.
[12]余胜,陈敬东,王新余.基于深度学习的复杂场景下车辆识别方法[J].计算机与数字工程,2018,46(9):1871-1875,1915.
[13]SKODRAS E.Rear lights vehicle detection for collision avoidance[C]∥IWSSIP 2012.IEEE,2012:134-137.
[14]CABANI I,TOULMINET G,BENSRHAIR A.Color-based detection of vehicle lights[C]∥Intelligent Vehicles Symposium.IEEE,2005:278-283.
[15]刘尊洋,叶庆,李菲,等.基于亮度与颜色四阈值的尾灯检测算法[J].计算机工程,2010,36(21):202-203,206.
[16]刘勃,周荷琴,魏铭旭.基于颜色和运动信息的夜间车辆检测方法[J].中国图象图形学报,2005(2):187-191.
[17]汽车及挂车转向信号灯配光性能:GB17509-2008[S].北京:中国标准出版社,2008.
[18]范红武,韩柯,孙亮,等.基于视觉自主车的尾灯灯语识别方法研究[J].电脑知识与技术,2010(12):9790-9792.

相关文章 15

[1]	王栋, 周大可, 黄有达, 杨欣. 基于多尺度多粒度特征的行人重识别 Multi-scale Multi-granularity Feature for Pedestrian Re-identification 计算机科学, 2021, 48(7): 238-244. https://doi.org/10.11896/jsjkx.200600043
[2]	张琳娜,陈建强,陈晓玲,岑翼刚,阚世超. 面向行车视频目标实时检测的轻量级SSD网络 Lightweight SSD Network for Real-time Object Detection in Automotive Videos 计算机科学, 2019, 46(7): 233-237. https://doi.org/10.11896/j.issn.1002-137X.2019.07.035
[3]	韩克堃, 胡桂川, 任静, 何鸿宇, 刘佳音. 图像处理在风电叶片法兰端面特征尺寸检测中的应用 Application of Image Processing in Feature Size Detection of Wind Turbine Blade’s Flange Face 计算机科学, 2019, 46(6A): 562-565.
[4]	韩绍超,徐遵义,尹中川,王俊雪. 指针式仪表自动读数识别技术的研究现状与发展 Research Review and Development for Automatic Reading Recognition Technology of Pointer Instruments 计算机科学, 2018, 45(6A): 54-57.
[5]	张文勇, 陈乐柱. 基于LabVIEW机器视觉的餐具分拣系统 Tableware Sorting System Based on LabVIEW Machine Vision 计算机科学, 2018, 45(6A): 595-597.
[6]	史凯静,鲍泓. 基于改进的FAST R-CNN的前方车辆检测研究 Forward Vehicle Detection Research Based on Improved FAST R-CNN Network 计算机科学, 2018, 45(6A): 179-182.
[7]	钱基德, 陈斌, 钱基业, 赵恒军, 陈刚. 基于机器视觉的液晶屏Mura缺陷检测方法 Machine Vision Based Inspection Method of Mura Defect for LCD 计算机科学, 2018, 45(6): 296-300. https://doi.org/10.11896/j.issn.1002-137X.2018.06.052
[8]	陈涵深, 姚明海, 陈志浩, 杨圳. 基于多帧叠加和窗口搜索的快速车道检测 Efficient Method of Lane Detection Based on Multi-frame Blending and Windows Searching 计算机科学, 2018, 45(10): 255-260. https://doi.org/10.11896／j.issn.1002-137X.2018.10.047
[9]	朱航江,朱帆,潘振福,朱永利. 运动状态与尺度估计的核相关目标跟踪方法 Visual Object Tracking Method with Motion Estimation and Scale Estimation 计算机科学, 2017, 44(Z11): 193-198. https://doi.org/10.11896/j.issn.1002-137X.2017.11A.040
[10]	董天阳,朱浩楠,王浩. 基于视频的夜间车辆检测与跟踪算法研究 Video-based Nighttime Vehicle Detection and Tracking Algorithm 计算机科学, 2017, 44(Z11): 233-237. https://doi.org/10.11896/j.issn.1002-137X.2017.11A.049
[11]	宣寒宇,刘宏哲,袁家政,李青,牛小宁. 一种鲁棒性的多车道线检测算法 Robust Multi-lane Detection Algorithm 计算机科学, 2017, 44(11): 305-313. https://doi.org/10.11896/j.issn.1002-137X.2017.11.047
[12]	李飞,张小洪,赵晨丘,鄢萌. 自适应的SILTP算法在运动车辆检测中的研究 Vehicle Detection Research Based on Adaptive SILTP Algorithm 计算机科学, 2016, 43(6): 294-297. https://doi.org/10.11896/j.issn.1002-137X.2016.06.058
[13]	徐邦振,汤一平,蔡国宁. 基于机器视觉的轻量级驾驶辅助系统 Machine Vision-based Lightweight Driver Assistance System 计算机科学, 2015, 42(Z11): 520-524.
[14]	徐振驰,纪磊,刘晓荣,周晓佳. 基于显著性特征的食用菌中杂质检测 Recognition of Impurities Based on their Distinguishing Feature in Mushrooms 计算机科学, 2015, 42(Z11): 203-205.
[15]	郭庆慧,梁秀霞,张锐. 基于稀疏特征点的单视点深度图像校准 Calibration for Single Viewpoint Depth Image Based on Sparse Feature Points 计算机科学, 2014, 41(Z6): 191-195.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed