透過您的圖書館登入
IP:3.16.66.179
  • 會議論文
  • OpenAccess

Fast Depth Coding of 3D-HEVC Based on Visual Perception

並列摘要


The standard for 3D video coding (3D-HEVC) has been established to enrich the multimedia experience and application. 3D-HEVC can make use of inter-view and inter-component correlations in the coding process. However, it increases the coding complexity which is not beneficial to real-time application. The goal of this paper is to reduce the depth coding complexity of 3D-HEVC encoder by utilizing the properties of visual perception. This paper proposes the vision-oriented edge detection for depth map by combining Sample Adaptive Offset(SAO) and Just Noticeable Depth Difference (JNDD) model. By analyzing depth map, the coding tree unit (CTU) is classified into various types and the corresponding mode decision scheme is proposed. The experimental results show that the proposed algorithm can provide much more time-saving than previous work and maintain the visual quality.

並列關鍵字

3D-HEVC depth coding fast algorithm JNDD SAO

延伸閱讀