Abstract
Defect detection plays a critical role in thin film transistor liquid crystal display (TFT-LCD) manufacturing. This paper proposes an inline defect-detection (IDD) system, by which the defects can be automatically detected in a TFT array process. The IDD system is composed of three stages: the image preprocessing, the appearance-based classification and the decision-making stages. In the first stage, the pixels can be segmented from an input image based on the designed pixel segmentation method. The pixels are then sent into the appearance-based classification stage for defect and non-defect classification. Two novel methods are embedded in this stage: the locally linear embedding (LLE) and the support vector data description (SVDD). LLE is able to substantially reduce the dimensions of the input pixels by manifold learning and SVDD is able to effectively discriminate the normal pixels from the defective ones with a hypersphere by one-class classification. After aggregating the classification results, the third stage outputs the final detection result. Experimental results, carried out on real images provided by a LCD manufacturer, show that the IDD system can not only achieve a high defect-detection rate of over 98%, but also accomplish the task of inline defect detection within 4 s for one input image.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
