LCD Mura Detection Based on Accumulated Differences and Multi-resolution Background Subtraction

A mura detection approach based on the difference accumulation and background estimation is proposed to detect mura in the thin film transistor liquid crystal display (TFT-LCD) images. We were motivated by wavelet transform to derive a multi-resolution method for accelerating the detecting speed of background estimation. While the visibility of mura depends on inspecting angles and online environmental light luminance, sometimes the defects may be invisible in the observed images. Thus, we calibrate the non-uniform luminance and accumulate the continuous online images for depressing noise and enhancing the contrast between mura and background. However, the online moving speed is unstable; thus, we propose a full-automatic non-uniform displacement calibration method. We demonstrate them by applying to real and artificial online image sequences. The efficiency and effect is described in experiments. Our methods are also compared with other mura detecting methods.