A cascaded Mura defect detection method based on mean shift and level set algorithm for active‐matrix OLED display panel

Abstract Active‐matrix OLED (AMOLED), as the next‐generation display technology, is being commercially promoted rapidly. And Mura defects occur unavoidable during different phases of the AMOLED panel production process. In this paper, a cascaded Mura detection method leveraging the mean shift and the level set algorithm is proposed. First, we use the mean shift algorithm to find the general contour of the Mura defects that circumvent the issue in established level set segmentation approach wherein the use of the local image information is sensitive to the initial contour. Then, we improve the level set model that combines global and local information to segment Mura defects accurately. The integration of local image information can levitate the challenges in the global image model, which cannot separate the local intensity inhomogeneity and texture background individually. The experiments show that the cascaded method has a superior capability in terms of both accuracy and efficiency.