A robust image blind watermarking algorithm based on adaptive quantization step in dwt 域自適應(yīng)量化步長(zhǎng)的圖像盲水印算法
The algorithm is based on the optimum quantization step for each dwt coefficient according to psychovisual considerations which derived from methods in image compression 本文的算法基于圖像壓縮算法中利用心理視覺(jué)給出的dwt系數(shù)的最優(yōu)量化步驟。
This paper can put into four parts ? this first part is the description and theoretical analyses of source coding , which focus on the research of optimizing equal quantization ? the second part presents the theoretical description of joint channel - source coding , which focus on the research of combined channel - source coding o the third part is about the application of combined channel - source coding to two different channel models , binary symmetric channel and cdma channel o in this part , two different coding designs are given according to different characters of these two channels ? and the last part is the description of simulation of combined channel - source coding ? most of my work are about two parts , one is to find the most appropriate quantization steps and centroid points of separate channel - source coding , another is to simulate the combined channel - source coding ? comparing the simulation results of separate channel - source coding and combined channel - source coding , the characters of joint channel - source coding are given 本論文可以分成四部分:第一部分給出了信源編碼的基本概念和理論分析,重點(diǎn)放在最優(yōu)均勻量化編碼的研究方面;第二部分給出了信道?信源聯(lián)合編碼的原理敘述,重點(diǎn)放在復(fù)合式信道?信源編碼的分析研究上;第三部分將信道-信源聯(lián)合編碼原理應(yīng)用在兩種噪聲信道上:離散無(wú)記憶信道和cdma信道,并根據(jù)兩種信道的不同特點(diǎn)詳細(xì)描述了兩種相應(yīng)的編碼設(shè)計(jì)方案;第四部分給出了復(fù)合式信道-信源編碼的仿真結(jié)果以及對(duì)結(jié)果的相應(yīng)分析。
As human eyes are more sensitive to the low coefficients than to the high coefficients , a preprocessor is employed to remove part of the high coefficients and a visual property factor is introduced to increase the quantization step in complex regions while decreasing the quantization step in flat regions 該策略利用人眼對(duì)圖像低頻分量比對(duì)高頻分量敏感的特性,先對(duì)圖像進(jìn)行預(yù)處理,去掉部分高頻分量,并在量化時(shí)引入視覺(jué)特性因子以加大圖像復(fù)雜區(qū)的量化而減小圖像平坦區(qū)的量化。
The adaptation processing includes linear prediction coefficient adaptation and adaptation of quantization step size for residual signals . based on g . 726 , we adopt a huffman coder to make use of probability statistic of bit cascade covering n ( n 1 ) samples generated from adpcm , in order to further reduce the bit rate . ng is lossless entropy coding , the speech quality of our improved algorithm should be same as that of g . 726 standard 我們的研究和改進(jìn)工作包括:研究最優(yōu)非均勻自適應(yīng)量化器,及其自適應(yīng)算法;研究波形預(yù)測(cè)函數(shù),以及函數(shù)零點(diǎn)、極點(diǎn)的自適應(yīng)算法;基于每n ( n 1 )個(gè)樣本所對(duì)應(yīng)符號(hào)的概率統(tǒng)計(jì),對(duì)預(yù)測(cè)殘差量化值再進(jìn)行huffman編碼,進(jìn)一步降低比特率。