Removal of ring artifacts resulting from digital holography system imperfections represent a challenging task. Reconstructed images from such systems are used in applications for pollen concentration monitoring, thus a reliable preprocessing of those images is needed. In this paper, the graph signal processing framework is proposed for suppression of ring artifacts by utilizing two graph filters: Taubin filter and optimal graph filter. Both designed filters are applied on reconstructed holographic images and their performances are evaluated through decreasing of total variation metric.