摘要
针对当前手机屏幕划痕检测存在的检测精度低,主观性强等问题,提出了一套定量检测方法。通过边缘检测算法得到划痕区域轮廓,在测量划痕区域像素点累加求出划痕面积;采用Rosenfeld细化算法求出弧长。实验表明,系统对划痕区域面积和弧长的检测精确度均可达98%以上。
In view of the low detection precision and subjectivity in the current screen scratch detection of mobile phone screen,a quantitative method is proposed. Getting the contour of the scratched region by the edge detection algorithm.The area scratched is measured by adding pixels.The Rosenfeld thinning algorithm is used to obtain the scratched skeleton and then the arc length is obtained. The experimental results showed that the detection accuracy of the area and arc length of the scratched area can be more than 98%.
引文
[1]易松松.基于机械视觉的手机面板缺陷检测方法研究[D].哈尔滨:哈尔滨工业大学,2016.
[2]张岩,王宝光.手机面板表面质量在线检测系统的研究[J].传感器与微系统,2010,29(4):52-54.
[3]刘泽,王嵬,王平.铁轨表面缺陷检测机器视觉系统的设计[J].电子测量与仪器学报,2010,24(11):1012-1017.
[4]苏孝雨,张荣辉,张杰.基于机器视觉的液晶屏裂痕自动检测方法研究[J].微型机与应用,2016,35(10):52-54.
[5]赵健,高军,罗超.基于数字图像处理的玻璃缺陷在线检测系统[J].测控技术与仪器仪表,2013,39(12):90-92.