查詢結果分析
相關文獻
頁籤選單縮合
題 名 | Gray Prediction Model Based Finite State Vector Quantization of Images |
---|---|
作 者 | Chang, Hsuan T.; Pan, Yann H.; Chang, Ting C.; Chen, Yuan B.; | 書刊名 | International Journal of Electrical Engineering |
卷 期 | 12:2 民94.05 |
頁 次 | 頁169-175 |
分類號 | 312.1 |
關鍵詞 | Gray theory; Finite-state vector quantization; Gray prediction model; Next-state function; |
語 文 | 英文(English) |
英文摘要 | Gray theory recently has received a great deal of attention because it has been successfully applied to many disciplines. In this paper, a new class of finite-state vector quantizers (FSVQs) whose next-state function design is based on the gray prediction model (GPM), named the GPMVQs, is proposed. In GPMVQs, the GPM with a single variable and a first-order differential equation is used to design the next-state function for constructing the state codebooks in PSVQs. Four pixels in each row or column of previously coded blocks are the inputs of the GPM and then the corresponding output predicts a pixel in the border of current block. With the western and northern blocks of the current block, seven border pixels are predicted and then used to choose the codewords in the super codebook to construct the state codebook. Simulation results show that the proposed GPMVQs can outperform the side-match vector quantizers, especially for less-complicated images. |
本系統中英文摘要資訊取自各篇刊載內容。