No abstract

In this paper, a n image coding scheme using t h e Discrete Cosine Transform is analyzed when t h e transform coefficients a r e vector quantized. The coding method is based on t h e known scheme proposed by W. Chen which sorts t h e picture blocks into classes according t o t h e level of image activity. The coding scheme is modified t o allow for vector quantization of t h e ac coefficients, in particular a Pyramid Vector Quantizer (PVQ) is used. This is based on t h e statistical and geometric properties of a Laplacian source which, in fact, is t h e best model for t h e ac coefficients of t h e two-dimensional Discrete Cosine Transform (2D-DCT) of an image. A method for forming almost statistically independent vectors is also suggested and improves quantization performance. Images are encoded with both t h e P V Q and standard scalar quantizer transform coders, demonstrating t h a t t h e P V Q coder reduces t h e mean square encoding error and improves image quality. In particular, emphasis is given t o how t h e use of fractional bit rates affects t h e objective and subjective gains obtained. The results presented (Le. mean square error values and printed images) have been obtained experimentally, working with a statistical criterion in a group of images whose size was in accordance with t h e 50 Hz C C I B Recommendation 6 0 1 Standard.

No abstract

No abstract

No abstract