Optimal Quantization Parameter Set for MPEG-4 Bit-Rate Control ∗

SUMMARY In this paper, we propose an optimal bit rate control algorithm which is fully compatible with MPEG-4 or H.263+. The proposed algorithm is designed to identify the optimal quantizer set through Lagrangian optimization when used for optimal bit allocation. To find the optimal quantizer set, we make use of the Viterbi algorithm in order to solve the dependency between quantization parameters of each macroblock due to the unique characteristics of MPEG-4 or H.263+. We set the Lagrangian cost function as a cost function of the Viterbi algorithm. We implement the proposed algorithm in MPEG-4 coders and compare its performance to the VM8 and optimal bit rate control algorithm, using independent quantization parameters in the circumstance of a low bit rate.

[1]  Antonio Ortega,et al.  Bit-rate control using piecewise approximated rate-distortion characteristics , 1998, IEEE Trans. Circuits Syst. Video Technol..

[2]  Antonio Ortega,et al.  Bit allocation for dependent quantization with applications to multiresolution and MPEG video coders , 1994, IEEE Trans. Image Process..

[3]  Jordi Ribas-Corbera,et al.  A frame-layer bit allocation for H.263+ , 2000, IEEE Trans. Circuits Syst. Video Technol..

[4]  Yair Shoham,et al.  Efficient bit allocation for an arbitrary set of quantizers [speech coding] , 1988, IEEE Trans. Acoust. Speech Signal Process..

[5]  Antonio Ortega,et al.  Optimal trellis-based buffered compression and fast approximations , 1994, IEEE Trans. Image Process..

[6]  K. Rijkse,et al.  H.263: video coding for low-bit-rate communication , 1996, IEEE Commun. Mag..

[7]  Jordi Ribas-Corbera,et al.  Rate control in DCT video coding for low-delay communications , 1999, IEEE Trans. Circuits Syst. Video Technol..