Technologies Research Of AOI For ITO Line Of Touch Screen

With the mobile Internet era, smart phones become an everyday tool for the most important connection. Touch screen is the basic components of smart phones. Its quality affects directly the performance and user feeling of smart phones. The complex preparation process of touch screen brings all kinds of defects. So the detection of the touch screen quality has become an important part of the production process. Most manufacturers use the manual sampling mode at present. Of course, the efficiency of artificial detection is low and miss rate is high. The application of automatic optical inspection(AOI) technology to the touchscreen ITO line defect detection has been a popular research direction. It includes optical, electrical, software and calculate interdisciplinary knowledge. For these reasons, it has a very high research value and a wide market value in inspection industry.In this dissertation,ITO defects AOI system design,defect feature extraction and defect identification classification and other key technologies have been researched. The main tasks in this dissertation are as follows.(1) First, according to the overall design scheme and the associated technical requirements of the system, the design and selection of the optical imaging system and motion control have been completed. In the motion mode of XY-axis, high-precision linear motor with high-speed linear CCD complete to collect the image of high quality.(2) Second, in the image preprocessing, the ITO line image of touch screen goes through image de-noising, two value segment, morphology processing. Especially, different results of processing algorithms are contrasted. And the method based on bimodal block iterative is proposed for the characteristics of the ITO line images. Target and background of image is been segment. It has a small amount of computation and high speed. A method based on projected changes using mark point positioning is adopted in terms of image registration. It has high speed and efficient registration. The ITO line texture features and defect geometry are researched. An improved ellipse fitting is proposed. It can exclude interfere of noise and get high fit. A tracking ITO line width template method based on geometric characteristics of defect is proposed in the defect classification.(3)Finally, C ++, Qt, OpenCV programming languages are used to complete the preparation of ITO detection system software. The information about the product data report has been designed. The three core modules of software are defect inspection, motion control and data report. The dissertation describes functions and interface effects.The experimental results show that the ITO line defect inspection system completes the requirements that the average detection time is within 35 s, the accuracy of defect recognition and classification rate is 90% for the size of 12×8 cm~2 touch screen. To our knowledge, it is the first report that AOI technology is applied to high-precision, highresolution ITO line of touch-screen defect inspection system which is real-time online in china.