针对7.62~20.3 cm (3～8 in)的手机液晶模组, 以机器视觉为基础, 图像处理为核心, 并集成视觉定位系统, 实现液晶模组各种类型缺陷的全自动化检测。使用定位视觉软件, 串联定位模块和运动控制模块工作, 精确控制模组连接器位置, 实现模组金手指的自动对接, 对准精度小于0.03 mm。使用面阵CCD采集液晶模组点亮后图像, 视觉检测软件提取ROI区域, 做Gabor滤波、双阈值二值化、图像去噪等预处理。采用二维直方图斜分法定位图形边缘区域进行模糊增强处理, 最后计算出缺陷属性, 缺陷检测次品识别率大于95%。所设计的液晶模组自动光学检测系统能够降低生产成本和退货率, 简化生产线应用, 对液晶模组组装产业开发缺陷检测系统具有一定参考价值。

(3.30英文摘要替换完了)The system is designed mainly for mobile LCD modules with a size of 7.62~20.3 cm (3～8 in) with image processing as the core based on machine vision. The visual positioning system is integrated to achieve the fully automatic detection of various types of module defects. Based on the cooperation of the positioning module and the motion control module, the positioning vision software is used to precisely control the position of the module connector and realizes the automatic docking of the connecting fingers with an alignment accuracy less than 0.03 mm. The array CCD is used to collect the images after the LCD module is lit up. The visual inspection software extractes the ROI area and performed preprocessing such as Gabor filtering, double threshold binarization and image denoising. The two-dimensional histogram oblique division method is used to locate the edge area of images for blur enhancement, and the defect properties is calculated. The defect detection has a defect recognition rate greater than 95%. The designed LCD module automatic optical inspection system can not only reduce the production costs and return rates, but also simplify the production line applications, provide a certain reference value for the development of defect detection systems for the LCD module assembly industry.