Adaptive Entropy Coded Subband Coding of Images

Recently. subband coding (SBC) has been the subject of rather extensive investigation in the area of image coding because of its potential advantages over fullband coding. In particular, it leads to subjectively more pleasing image reconstructions and allows progressive transmission. However, most of the work on SBC has been based on level constrained design of the embedded quantizers. In this paper, we describe a new design approach, called 2-D entropy-constrained subband coding (ECSBC), based upon recently developed 2-D entropy-constrained vector quantization (ECVQ) and 2-D entropy-constrained predictive vector quantization (ECPVQ) schemes. The output indexes of the embedded quantizers are further compressed by use of noiseless entropy coding schemes, such as Huffman or arithmetic codes, resulting in variable rate outputs. Depending upon the specific configurations of the ECVQ and the ECPVQ over the subbands, many different types of SBC schemes can be derived within the generic 2-D ECSBC framework. Among these, we concentrate on three representative types of 2-D ECSBC schemes and provide relative performance evaluations. In particular, one of the 2-D ECSBC schemes that employs a 2-D ECPVQ for the lowest frequency subband with separate entropy-constrained scalar quantizers (ECSQ’s) applied to the remaining higher frequency subbands is shown to provide excellent quality image reconstructions at average rates as low as 0.3 b/pixel for typical real-world images. We also describe an adaptive buffer instrumented version of 2-D ECSBC, called 2-D ECSBC/AEC, for use with fixed rate channels which completely eliminates buffer overflow/under-flow problems. This adaptive scheme achieves performance quite close to the corresponding ideal 2-D ECSBC system. We demonstrate results on a selected real-world image.