摘要
在工业生产中,对前照灯检测的精度和速度要求越来越高。在汽车前照灯检测原理的基础上,通过HALCON和Microsoft Visual Studio完成前照灯检测系统的研究设计。针对车灯的光形图像,通过亚像素坎尼边缘提取法、Tukey加权函数的线性拟合法、灰度重心法等,实现了近光明暗截止线、近光拐点、远近光光束中心及光轴偏转角等参数的准确检测。实验结果表明:该系统检测精度高,稳定性好,具有较高的应用价值。
The accuracy and speed of headlamp detection are getting higher and higher in its manufacturing process. Based on the headlamp detection principle,the research and design of headlamp detecting system using HALCON and Microsoft Visual Studio were completed. The system can measure the characteristic parameters accurately such as line between bright part and dim part,center of low beam and high beam,corner of low beam,angle of light from headlamp. Some image processing solutions on the light shape of headlamp include Canny edge extraction based on subpixel precision,fitting line through the Tukey weight function,gray center function etc. The experimental results show that the system has high detection accuracy,good stability and high application value.
引文
[1]张勇.基于图像处理的汽车前照灯检测系统的研究[D].长春:长春理工大学,2005.
[2]倪高平.汽车前照灯光型检测系统[D].苏州:苏州大学,2015.
[3]张志瀚.基于机器视觉的汽车前照灯检测系统研究[D].武汉:武汉理工大学,2015.
[4]杨少荣.机器视觉算法与应用[M].北京:清华大学出版社,2008.
[5]崔明,顾启民.基于机器视觉的前照灯拐点检测与研究[J].制造业自动化,2015,37(10):13-16.CUI M,GU Q M. Detection and research for the crosspoint of headlamp based on machine vision[J]. Manufacturing Automation,2015,37(10):13-16.
[6]王文成.基于Halcon的齿轮缺陷检测系统设计[J].机械传动,2014,38(9):60-63.WANG W C.Design of gear detection system based on the Halcon[J].Journal of Mechanical Transmission,2014,38(9):60-63.