Please use this identifier to cite or link to this item:
標題: 以離散小波轉換及SPIHT為基礎進行即時NTSC影像壓縮之編碼器
作者: 劉洽興
關鍵字: 離散小波轉換;SPIHT;影像壓縮
出版社: 電機工程學系
在現今的人類生活中, 影像壓縮變得越來越重要,舉凡影像監視器、視訊會議、VCD錄影等日常生活常遇到的物品,皆有影像壓縮的技術在其中發揮作用。因為影像的資料量非常大,如果不經壓縮就直接進行通訊和儲存,那將耗費大量的頻寬和儲存空間。
本論文主要討論利用FPGA來設計實現視訊信號壓縮的用途。因此使用Haar離散小波轉換將原始資料分離成不同頻域之資料型態,再經過有別於標準之SPIHT編碼方式,以Listless 和分割頻域之SPIHT方式將資料連成資料串,以達到高壓縮比之即時影像壓縮。其中尚需要搭配以樹為基礎之Scan order、DC和AC 分離、直接減去法及分割頻域等方式將暫存器的使用量降低至FPGA可接受的範圍,並且把資料的處理能力提高至每一秒壓縮640 X 240 的影像達15個畫面。

In our daily life, image compression has become more and more important. For example, the video monitor, video conferencing and VCD recorder etc., include the image compression technology. Since the data size of the image is very large, if there is no compression before transmitting or storing, it will waste a lot of communication bandwidth or storage space.
In this thesis, FPGA implementation of real-time video compression is described. The raw data are transformed to the multi-band data by Haar discrete wavelets transform. Because we want to get high compress ratio for real-time video, the multi-band data are encoded to series data by listless and multi-part SPIHT. Four methods are used to reduce the number of registers to an acceptable number for FPGA, i.e., the scan order based on tree, encoding DC and AC by different methods, direct eliminating multi-band data instead of refining, pre-encoding high-band and low-band data in different logic. The data processing speed for compressing a 640 x 240 picture can reach 15 frames per second.
Appears in Collections:電機工程學系所

Show full item record
TAIR Related Article

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.