查詢結果分析
來源資料
頁籤選單縮合
題名 | 層階式組合編碼在醫學影像之表現=The Performance of Hierarchical Combined Coding for Medical Image |
---|---|
作者 | 蘇振隆; 顏陳河; 許世勳; 賴麗娟; | 書刊名 | 中原學報 |
卷期 | 26:1 1998.02[民87.02] |
頁次 | 頁83-91 |
分類號 | 410.1644 |
關鍵詞 | 層階式組合編碼; 位元平面; 資料重複多餘性; 可復原壓縮; 不可復原壓縮; Hierarchical combined coding; Bit plane; Data redundancy; Reversible compression; Irreversible compression; |
語文 | 中文(Chinese) |
中文摘要 | 影像資料之壓縮對於資料庫的負擔及網路的傳輸扮演重要角色。本論文提出一種 層階式組合編碼(Hierarchical Combined Coding)方法,以位元平面觀點將醫學影像分割 成高位元平面影像(HBPI)及低位元平面影像(LBPI)。此法可提供可復原壓縮及不可復原 壓縮。HBPI採用無失真(可復原)壓縮的方式,而LBPI的壓縮則有選擇性。在不可復原壓縮 方面,由於HBPI包含影像中大部分主要的資訊並且採用無失真之壓縮,因此能在復原影像失 真較少的情況下提供較高的壓縮比。共用30張分別由X-光、電腦斷層攝影及磁振造影所形 成之醫學影像進行系統測試。研究結果中,將含八位元平面之醫學影像分成HBPI佔一位元、 LBPI佔七位元時不論在可復原壓縮方面或不可復原壓縮方面皆能提供最佳的壓縮表現。在可 復原壓縮方面,本研究壓縮法之壓縮比較商用壓縮法平均有6.57%之提昇。在不可復原壓縮 方面,壓縮比同為11之條件下,本研究壓縮法之PSNR較JPEG壓縮平均有11.52%之提昇。此外 針對醫學影像特性,消除不必要之影像背景雜訊,對提高壓縮比亦有幫助。 |
英文摘要 | Compressing medical images can save image storage spaces and transmission time. A technique is presented to compress medical images employing a hierarchical combined coding (HCC) method. Both lossy (irreversible) and lossless (reversible) implmentation are considered. The original image is first resolved into two kinds of bit plane images which have different stochastic properties. The high bit plane images (HBPI) which contained the most significant information was coded with lossless implmentation. The low bit plane images (LBPI) could be chosen to be coded with either lossless or lossy implmentation. Thirty images which generated from X-ray, Computized Tomography (CT), and Magnetic Resonance Imaging (MRI) were used for testing the system. The results showed that a combination of HBPI with one bit and LBPI with seven bits for eight bits plane medical images presented best performance in both lossy and lossless cases. Compared with commercial product, the lossless compression can improve compress ratio about 6.57%. By comparing with the JPEG standard, this method could increase PSNR to about 11.52% with the same compression ratio at 11 in lossy compression case. In addition, medical images with erasing background noise are very helpful in increasing compression ratio. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。