基于机器视觉的手机屏幕缺陷检测方法研究

Research on Detecting Method of Mobile Phone Screen Defect Based on Machine Vision

由于手机屏幕缺陷具有多样性和复杂性,几乎不可能存在一种通用的检测算法可以检测出手机屏幕可能存在的所有缺陷。笔者针对手机屏幕常见的坏点、坏线、色斑、mura和漏光等缺陷检测方法进行研究,开发了一系列的自动检测算法。可根据特定需求,选择单个或多个检测算法对工业相机采集的手机屏幕图像进行实时检测,判断手机屏幕是否符合出厂标准,并及时、准确地定位出屏幕缺陷区域。

Due to the diversity and complexity of mobile phone screen defects, it is almost impossible to have a universal detection algorithm that can detect all possible defects of the mobile phone screen. In this paper, we have developed a series of automatic detection algorithms for common defects. According to specific requirements, you can choose one or more detection algorithms to detect the screen image in real time, to determine whether the screen of the mobile phone meets the factory standards, and timely and accurately locate defects in the screen area.