Bit allocation and rate control algorithm for MVC

Abstract

Since the current multi-view video coding (MVC) software does not contain any rate control technique, this paper proposes a rate control algorithm for MVC based on the quadratic rate-distortion (R–D) model. The proposed algorithm classifies each picture into six frame types based on the relation between disparity prediction and temporal prediction estimation and also improves the estimation accuracy of the mean absolute difference (MAD). The proposed method allocates the bits and controls the rate for inter-view, frame layer and basic unit layer based on the analysis of the previously coded information. Compared to the multi-view video coding with fixed quantisation parameter, the proposed scheme achieves up to 0·25 dB improvements in peak signal-to-noise ratio (PSNR). Meanwhile, it can efficiently control the bit-rate with an average rate control error of 0·54%.

Keywords:

• multi-view video coding
• rate control
• bit-allocation
• rate–distortion model
• correlation analysis

This work is supported by the Natural Science Foundation of China (grant nos. 60832003, 60902085 and 60972137) and the Key Project of Shanghai Education Committee (09ZZ90). This work is also sponsored by the Natural Science Foundation of Shanghai (no. 09ZR1412500) and Innovation Foundation of Shanghai (nos. 10YZ09 and SHUCX091061). We thank the Interactive Visual Media Group of MERL, KDDI and Nagoya University/Tanimoto Lab for the data we used.