Guided preprocessing based on visual perception for multiview coding

Video processing considering human visual perception attracts more attention in recent years. Based on characteristic of visual perception, a guided preprocessing method for multiview coding is proposed in this paper. Texture preprocessing is fulfilled by Gaussian filtering on the background, which led by the mask determined by depth map. And guided filtering is adopted for depth preprocessing, where guidance map is generated by luminance and chrominance components from both depth map and texture map. The resultant texture video and depth video are applied for multivew coding finally. Subjective evaluation towards the resultant video proves that the visual quality obtained by the proposed method is better than the anchor. As well as for objective evaluation in terms of PSNR index, the proposed method is superior to the anchor.