Automatic Inspection for LCD Spacer

• Main Authors: Pauli Chen, 陳銘仁
• Other Authors: J. H. Jeng
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/17090603920130858802

碩士 === 義守大學 === 資訊工程學系 === 92 === There are many categories for LCD used in electric and communication products. All of the manufacturers are trying to make mass and reliable products. In this study，we propose an automatic system to detect spacers and check the quality of LCD according to the amount of spacers. Sizes of the spacers are generally very small and dispersed in the LCD panel. Detecting the amount of spacers by human eyes is inefficient and un-precise. Automatic detection of LCD spacers based on image processing is proposed in this thesis. After Image processing，we count the amount of spacers and judge the quality of LCD. The gray histogram is first constructed based on YCbCr color space and smoothed by an average filter. Then a threshold is determined to segment the gray level image to a binary image. The segmentation results are poor when the image are partially highlighted or shaded in a few of uniform regions. We here propose a method using histogram thresholding based on the gray histogram smoothed by an average of seven points and the differentiation histogram by an average of five points. The shapes of spacers are then extracted by mathematical morphology method. Finally，we use chain code to obtain the amount of spacers. The performance of the image-based method is proved to be better that done by human eyes.