Font Size: a A A

Lcd Defect Automatic Detection System

Posted on:2009-03-06Degree:MasterType:Thesis
Country:ChinaCandidate:G ZhangFull Text:PDF
GTID:2208360245978617Subject:Measuring and Testing Technology and Instruments
Abstract/Summary:PDF Full Text Request
At the present time, LCD has been the major part of display device and gradually taken the place of LED and VFD in onboard display system. Defect may occur because of the process or the circumstance in production flow, and now, inspections are made mainly by human inspector. For improving the efficiency and precision, it's very necessary to develop automatic inspection system for LCD based on machine vision.This article does some relational research work focus on developing automatic inspection system for LCD and introduce the typical defects (single pixel and line pixel defects) with their characteristics and origins. For each kind of defect, corresponding measurement and dealing method has been proposed, and then, the experimental hardware of the system has been made up. Thoroughly theoretical and experimental research has been done about the defect identify and automatic inspection based on image-processing.First, the resolution of CCD cannot identify the defect in one image by full screen, so, we separate the LCD into several area to solve the problem perfectly. Next, by using several methods of image disposal and analysis such as picture-segmentation, edge pick-up, mathematics morphology to deal with collected images , we get a better balance between speed and precision. Finally, we build an experimental platform after scientific reasoning and get a satisfactory result by plenty of experiments, make it a reality preliminary for automatic checkout system.The result proves that we can make it a reality that automatic inspecting the typical defects of LCD and prepare ourselves for the next actual application ahead. At the end, with the actual experience in researching work, some insufficiency and thinking of improvement about the system are also pointed out.
Keywords/Search Tags:machine vision, defect inspection, image processing, pattern identify
PDF Full Text Request
Related items