Image coding based on selective quantization of the reconstruction noise in the dominant sub-band.

Previous research has shown the benefits of using sub-band coding and vector quantization (SBC-VQ) for digitizing image frames at rates below 1 bit per pixel. An important factor in SBC-VQ design is the coding of the dominant sub-bands. These sub-bands typically need to be quantized with a resolution several times greater than the average bit rate. Since vector quantizers are particularly suited for low bit rate designs, special techniques are needed to accomplish high bit rate VQ coding of these sub-bands. In the work reported here, multi- stage quantization is used to encode the intensity waveform in the most dominant sub-band. The reconstruction error in the first stage of the process is itself quantized and the first-stage reconstruction is refined by removing from it the quantized reconstruction errors. The reconstruction error is coded on the basis of a bit allocation algorithm that utilizes the fact that perceptually significant error samples tend to occur in clusters in high-entropy parts of the input image frame.