An Efficient Vision Algorithm for fast and fine Mask-Panel Alignment

In a display or semiconductor mass production line, it is highly required to make an advance for the alignment process technology to reduce the overall tact time. If the vision that extracts the misalignment quantity between mask and panel is progressed, there are a lot of chances to achieve higher productivity. In this paper, based on these underlying assumptions to reduce the tact time, we focus on upgrading the level of vision algorithm which determines the misalignment between mask and panel. First of all, we are to investigate existing algorithms and propose an effective machine vision algorithm for the alignment system. Then, by using the visual alignment experimental setup composed of CCD camera and control PC, we verify how well it performs