Bit-plane-based lossless depth-map coding
Kyung Yong Kim · Optical Engineering · 2010
This work proposes an efficient bit-plane-based lossless depth-map coding method for an MPEG 3-D video coding scheme. This method uses the distinctive image characteristics between bit planes of the depth map. In the simulations, the performances of the proposed coding method are compared with those of the conventional lossless coding methods, i.e., JPEG-LS, JPEG-2000, and H.264/AVC, in terms of the intra- and also intercoding modes. In intracoding mode, the proposed method achieves the highest compression ratios as 14.28:1 on average. JPEG-LS, JPEG2000, H.264/AVC (CAVLC), and H.264/AVC (CABAC) obtain the compression ratios as 9.74:1, 7.68:1, 9.13:1, and 10.97:1, respectively. In intercoding mode, the proposed method also accomplishes the highest compression ratios as 28.91:1 on average. However, H.264/AVC (CAVLC) and H.264/AVC (CABAC) obtain the compression ratios as 19.82:1 and 23.45:1, respectively.