Automatic detection of defective shadow mask holes using the periodicity

With the production of high resolution TFT-LCD display, the visual inspection of its shadow mask is very important to its production process for the quality assurance. Since the size of LCD becomes large, the visual inspection of shadow mask cell is very time-consuming. This paper presents a new automated inspection method of the individual cell of LCD shadow mask, which can detect the very small cell defects reliably and rapidly. In fact, the line image data of shadow mask cells are periodically placed. The inspection method mainly depends on the fact that a small defect can be significantly emphasized in terms of the comparison between image pixels departed by the period. This comparison process is implemented as the digital filtering algorithm, which can be rapidly processed by VLSI digital signal processing chip. In addition to the individual hole defect, the average intensity of the shadow mask holes are also inspected according to the specification data. From the experimental results, this scheme shows a reliable inspection capability for individual cell defects.