الفهرس Only 14 pages are availabe for public view
 |  | from | 82 |  |  |
| | | from | 82 | | |
Abstract
Vector Quantization (VQ) techniques are known to be effective for signal compression, image processing, and pattern recognition [1],[7],[4]. The Vector Quantizer consists of a codebook and a search algorithm. The codebook design finds the best representative vectors for a training set vector of similar expected images. The designed codebook is used by both the encoder and the decoder. In VQ systems the encoder determines the closest codevector in the codebook against an incoming vector using a search algorithm and thereby mapping the input vectors into a set of index numbers. These index numbers are then transmitted through a channel. At the receiver, the decoder replaces the input vectors by the indexed codevectors and reconstructs the image. Cost of transmitting an image represented by the bit rate is decided by the number of the codevectors in the codebook n and the quality depends on the goodness of the codebook. The basic problem in the VQ techniques lies in how to generate a codebook that achieves good performance with low bit rate.
In this research, a new fast one-step codebook design algorithm for vector quantization is proposed for image coding. The method utilizes fuzzy classification scheme for best clustering of data vectors in the training space. The codebook design process terminates in just one step so that it is highly computationally efficient as compared to other reported methods. To improve the coding efficiency, image blocks are classified into two classes and a different codebook is designed for each class. The two codebooks are augmented to form one codebook so that there is no need to send class information. Moreover, entropy coding is used to send codevector index to the receiver for further bit rate reduction. Simulation results are presented to show the superior performance of the proposed method in terms of PSNR as compared to the state-of-the-art codebook design algorithms.