Premium
A new shape‐vector quantization‐based adaptive predictive image coder
Author(s) -
Wang Jian,
Naghdy Golshah
Publication year - 1999
Publication title -
international journal of imaging systems and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.359
H-Index - 47
eISSN - 1098-1098
pISSN - 0899-9457
DOI - 10.1002/(sici)1098-1098(1999)10:6<419::aid-ima2>3.0.co;2-5
Subject(s) - lossless compression , vector quantization , computer science , quantization (signal processing) , image compression , artificial intelligence , algorithm , pattern recognition (psychology) , data compression , pulse code modulation , pixel , coding (social sciences) , computational complexity theory , computer vision , image (mathematics) , mathematics , image processing , transmission (telecommunications) , telecommunications , statistics
In this paper, a new lossless image compression technique, shape‐vector quantization (VQ)‐based adaptive predictive coder (SAPC), is introduced. In the proposed scheme, the local shape information of the image block is obtained through shape‐VQ. This information is utilized by a novel predictive coder, shape‐differential pulse code modulation (DPCM), to adaptively select the optimum predictor on a pixel‐by‐pixel basis. The prediction errors can be further compressed by an error‐adjusting process. The proposed scheme achieves a breakthrough in prediction by utilizing the local feature of the image block through shape‐VQ, thus improving the accuracy of the prediction while reducing the overhead of the side information. It also simplifies the complicated procedures involved in the computation of the prediction parameters. Although the proposed scheme outperforms many traditional lossless image‐coding schemes, it produces comparable results to the newly developed context‐based scheme with lower computational complexity. On the basis of the promising compression results, the proposed scheme could be the best candidate for the lossless image coding. © 2000 John Wiley & Sons, Inc. Int J Imaging Syst Technol, 10, 419–426, 1999