An effective image segmentation method for noisy low-contrast unbalanced background in Mura defects using balanced discrete-cosine-transfer (BDCT)

► We present a new image segmentation method to detect the Mura defects. ► Mura detection is based on a balanced discrete-cosine-transfer (BDCT) method. ► We overcome the limitations of DCT methods in reconstructing image background. ► Six important steps are taken essentially to extract Mura defect regions.