Algorithms for Efficient and Fast 3D-HEVC Depth Map Encoding by Gustavo Sanchez & Luciano Agostini & César Marcon

Author:Gustavo Sanchez & Luciano Agostini & César Marcon
Language: eng
Format: epub
ISBN: 9783030259273
Publisher: Springer International Publishing

4.1.2 Block Sizes and Encoding Mode Distribution

Figure 4.3 shows the block size percentage selected to encode the depth maps according to the QP-depths. This experiment considers results for all videos inside CTC when evaluated under all-intra-frame configuration.

Fig. 4.3Block size distribution for four quantization scenarios and five block sizes

The selection of the block size is highly dependent on the QP-depth value; i.e., high QP-depth values imply selecting bigger blocks (e.g., 32 × 32 and 64 × 64 block sizes are used more than 75% of times for QP-depth = 45), and the opposite is also true (e.g., 4 × 4 and 8 × 8 block sizes are used more than 65% of times for QP-depth = 34). Besides, 16 × 16 blocks selection percentage remains nearly constant, i.e., independently of the QP-depth. This happens because low QP-depths tend to preserve the depth map details, generating heterogeneous regions, which are encoded better with smaller block sizes. On the other hand, high QP-depths tend to attenuate the depth map details, creating homogeneous regions, which are encoded better with bigger block sizes. Besides, it is crucial to understand the encoder decisions according to the selection of the available encoding flows. This analysis is presented next, considering five possibilities of encoding flows:Bp-TQ – a bipartition mode is selected as the final encoder decision using the TQ flow;

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.