查詢結果分析
來源資料
頁籤選單縮合
| 題 名 | A High Fidelity Image Coding Using VQ-BTC=採用VQ-BTC技術之高傳真度影像壓縮編碼 |
|---|---|
| 作 者 | 王周珍; 謝朝和; | 書刊名 | 中正嶺學報 |
| 卷 期 | 27:2 1999.01[民88.01] |
| 頁 次 | 頁103-119 |
| 分類號 | 312.1 |
| 關鍵詞 | 影像處理; 向量量化; 區塊截斷編碼; 訊號處理; Image processing; Vector quantization; Block truncation coding; Signal processing; |
| 語 文 | 英文(English) |
| 中文摘要 | 本文提出一個能結合向量量化和區塊截斷編碼各自優點之高傳真度影像編碼技術 。首先設計一分類器來判斷每一影像區塊亮度之對比性,低對比性區塊非常適合直接採用向 量量化來加以編碼,而高對比性區塊因含有邊線和紋路的資訊,為了使邊線能被完整保留, 一個利用區塊截斷編碼特性之訊號處理器被設計來將其分解為邊緣訊號和高相關性區塊,使 它們更適合採用向量量化編碼來達到最低的失真。和其他壓縮編碼技術比較起來,本方法在 低位元率下,更能保有重建影像邊線的完整性及傳真度。 |
| 英文摘要 | A high fidelity image coding technique which combines the advantages of both vector quantization (VQ) and block truncation coding (BTC) is presented in this paper. An image block is first classified into a low-contrast block or a high-contrast block. Each low-contrast block is encoded directly with VQ. Each high-contrast block is decomposed into the edge signals and highly correlated block, such that the edges of a block can be preserved very well and a highly correlated block can be more efficiently encoded by VQ with low distortion. Compared with the conventional BTC, the new technique provides an improvement in reconstructed edge with much lower bit rates. |
本系統中英文摘要資訊取自各篇刊載內容。