New bit allocation method for wavelet and octave subband coding

A new bit allocation algorithm is derived for wavelet and octave subband source coding, subject to the Shannon rate-distortion bound, using a Lagrange multiplier optimization technique. A sequential form is developed from the derived direct form, for effectively reducing unassigned bits due to negative bit allocation and fixed bit-rate quantizers. The proposed sequential form algorithm is generalized and optimized to incorporate any given finite number of fixed bit-rate quantizers and is highly suitable for real-time applications and implementation. The bit allocation performance of the sequential Shannon-bound based algorithm is demonstrated through its application in image coding.

[1]  M. Bernfeld,et al.  Optimum Rate Allocation for Encoding Sets of Analog Messages , 1965 .

[2]  Michel Barlaud,et al.  Image coding using wavelet transform , 1992, IEEE Trans. Image Process..

[3]  T. A. Ramstad Considerations on quantization and dynamic bit-allocation in subband coders , 1986, ICASSP '86. IEEE International Conference on Acoustics, Speech, and Signal Processing.

[4]  Sohail Zafar,et al.  Motion-compensated wavelet transform coding for color video compression , 1992, IEEE Trans. Circuits Syst. Video Technol..

[5]  P. Schultheiss,et al.  Block Quantization of Correlated Gaussian Random Variables , 1963 .

[6]  John W. Woods,et al.  Subband coding of images , 1986, IEEE Trans. Acoust. Speech Signal Process..

[7]  L. M. Goodman,et al.  Optimum rate allocation for encoding sets of analog messages , 1965 .

[8]  J. Woods,et al.  Sub-band coding of images , 1986, ICASSP '86. IEEE International Conference on Acoustics, Speech, and Signal Processing.