發(fā)布時(shí)間：2018-11-26 07:06

【摘要】：近年來，支持多級(jí)輸出的打印機(jī)涌現(xiàn)市場(chǎng)，它們通過高尋址技術(shù)或改變打印頭設(shè)計(jì)等方法來控制微小墨滴的落點(diǎn)，，從而在承印物上實(shí)現(xiàn)多級(jí)輸出。多級(jí)打印輸出需要有相應(yīng)的多級(jí)加網(wǎng)軟件模塊來支持，而多級(jí)加網(wǎng)算法是這一軟件模塊的核心。本研究通過分析現(xiàn)有的多級(jí)加網(wǎng)算法，發(fā)現(xiàn)它們存在算法效果不理想、復(fù)雜度太高或者對(duì)硬件要求過高等缺陷。其中最為人們普遍接受的多級(jí)誤差擴(kuò)散算法也存在著明顯的缺陷，即半色調(diào)后的圖像在中間調(diào)緩變區(qū)域存在偽輪廓現(xiàn)象。針對(duì)這一缺陷，本文引入數(shù)學(xué)形態(tài)學(xué)中的開運(yùn)算操作對(duì)多級(jí)誤差擴(kuò)散半色調(diào)圖像進(jìn)行處理，獲得偽輪廓區(qū)域圖像，并采用鄰域生長(zhǎng)和填充補(bǔ)償?shù)姆椒▽?duì)偽輪廓進(jìn)行修正。實(shí)驗(yàn)證明論文提出的新算法可以在一定程度上改善甚至消除偽輪廓現(xiàn)象。 本文主要研究?jī)?nèi)容包括以下幾點(diǎn)： (1)分析現(xiàn)有的二級(jí)半色調(diào)算法以及多級(jí)誤差擴(kuò)散算法，研究基于多級(jí)誤差擴(kuò)散半色調(diào)算法的圖像中出現(xiàn)“偽輪廓”的主要原因。 (2)利用數(shù)學(xué)形態(tài)學(xué)方法處理半色調(diào)后的圖像，查找“偽輪廓”出現(xiàn)位置，控制開運(yùn)算算子以獲得準(zhǔn)確的“偽輪廓”邊緣。 (3)利用Matlab仿真，采取“填充補(bǔ)償”以及“鄰域生長(zhǎng)”兩種方法分別對(duì)存在偽輪廓的半色調(diào)圖像進(jìn)行修正，并比較兩種算法的優(yōu)劣。 (4)采用Matlab——JAVA混合編程方法實(shí)現(xiàn)噴墨打印的多級(jí)調(diào)頻加網(wǎng)軟件模塊的開發(fā)，并借助ESC語言驅(qū)動(dòng)Epson7880噴墨打印機(jī)輸出多級(jí)圖像。 本文的創(chuàng)新點(diǎn)主要體現(xiàn)在： (1)思路創(chuàng)新，首次將數(shù)學(xué)形態(tài)學(xué)的方法引入圖像半色調(diào)過程中，用以查找多級(jí)半色調(diào)圖像中存在的偽輪廓區(qū)域。 (2)首次提出了采用鄰域生長(zhǎng)和填充補(bǔ)償兩種方法對(duì)偽輪廓區(qū)域進(jìn)行修正。 (3)在對(duì)半色調(diào)圖像進(jìn)行客觀評(píng)價(jià)時(shí)，嘗試性地提出用分塊的PSNR值作為評(píng)價(jià)參數(shù)，通過實(shí)驗(yàn)驗(yàn)證該評(píng)價(jià)參數(shù)是有效可靠的。 本文的修正算法可以減輕甚至完全消除原算法中的偽輪廓現(xiàn)象，本文提出的算法修正的思路也可供其它半色調(diào)圖像處理借鑒和參考。 本論文得到江蘇省研究生創(chuàng)新工程項(xiàng)目“噴墨印刷的色彩信息高保真再現(xiàn)研究”(CXZZ110517)的支持_。 論文的研究成果已經(jīng)申請(qǐng)專利，專利申請(qǐng)?zhí)枮椋?01210001099.8。
[Abstract]:In recent years, there has been a market for printers that support multistage output. They control the drop point of tiny ink droplets by means of high addressing technology or changing printing head design, so as to achieve multistage output on the substrate. The multilevel print output needs to be supported by the corresponding multilevel netting software module, and the multilevel netting algorithm is the core of the software module. Through the analysis of the existing multilevel netting algorithms, it is found that they have some shortcomings, such as poor performance, too high complexity or too high hardware requirements. The most widely accepted multilevel error diffusion algorithm also has obvious defects, that is, the half-tone image has pseudo-contour phenomenon in the middle slow region. In order to solve this problem, the open operation in mathematical morphology is introduced to process the multilevel error diffusion halftone image, and the pseudo contour region image is obtained, and the pseudo contour is corrected by neighborhood growth and filling compensation. Experimental results show that the proposed new algorithm can improve or even eliminate the false contour phenomenon to some extent. The main contents of this paper are as follows: (1) analyze the existing two-level halftone algorithm and multi-level error diffusion algorithm, and study the main reasons of "pseudo contour" in the image based on multilevel error diffusion halftone algorithm. (2) the half-tone image is processed by mathematical morphology method, the position of "pseudo contour" is found, and the operating arithmetic is controlled to obtain accurate "pseudo contour" edge. (3) the half-tone image with pseudo-contour is modified by Matlab simulation, and the advantages and disadvantages of the two algorithms are compared by using "fill compensation" method and "neighborhood growth" method. (4) the multilevel FM screening software module of inkjet printing is developed by using Matlab--JAVA mixed programming method, and the multilevel image output of Epson7880 inkjet printer is driven by ESC language. The main innovations of this paper are as follows: (1) the method of mathematical morphology is introduced into the process of image halftone for the first time to find the pseudo-contour region in multilevel halftone images. (2) neighborhood growth and filling compensation are used to modify the pseudo-contour region for the first time. (3) in the objective evaluation of halftone images, we try to use the PSNR value of blocks as the evaluation parameters, and the experimental results show that the evaluation parameters are effective and reliable. The modified algorithm in this paper can reduce or even eliminate the false contour phenomenon in the original algorithm. The proposed algorithm can also be used for reference in other halftone image processing. This thesis is supported by the innovative engineering project of Jiangsu Province, "High fidelity reproduction of color information in ink jet printing" (CXZZ110517). The research results of the thesis have been applied for patent, patent application number: 20121000 1099.8.
【學(xué)位授予單位】：南京林業(yè)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2012
【分類號(hào)】：TP334.8

【參考文獻(xiàn)】

相關(guān)期刊論文 前9條

1 趙小梅;陳駿驄;;數(shù)字加網(wǎng)圖像的質(zhì)量評(píng)價(jià)研究[J];包裝工程;2009年01期

2 王敏惟,黃進(jìn),陳其昌;熱印字機(jī)的數(shù)字半色調(diào)算法——多級(jí)多階誤差分散算法研究[J];計(jì)算機(jī)工程與應(yīng)用;2002年06期

3 周景超;戴汝為;肖柏華;;圖像質(zhì)量評(píng)價(jià)研究綜述[J];計(jì)算機(jī)科學(xué);2008年07期

4 趙德方;鄒鯤;;基于誤差擴(kuò)散的半色調(diào)圖像輸出算法修正[J];機(jī)械制造;2010年09期

5 劉小丹;楊燕;;一種大幅面二級(jí)設(shè)備輸出多級(jí)半色調(diào)圖像的方法[J];計(jì)算機(jī)應(yīng)用;2006年12期

6 劉小丹;蔡靜;;基于多級(jí)半色調(diào)技術(shù)的大幅面圖像二級(jí)多遍輸出方法[J];計(jì)算機(jī)科學(xué);2006年09期

7 任小玲,張二虎;數(shù)字半色調(diào)技術(shù)發(fā)展概述[J];西安工業(yè)學(xué)院學(xué)報(bào);2003年02期

8 趙小娥;;半色調(diào)圖像質(zhì)量評(píng)價(jià)研究[J];西安工程科技學(xué)院學(xué)報(bào);2006年03期

9 楊燕;劉小丹;;消除輸出畫面?zhèn)屋喞男滤惴╗J];中國(guó)科技信息;2008年09期

相關(guān)碩士學(xué)位論文 前5條

1 李育喜;調(diào)頻加網(wǎng)制版工藝研究[D];西安理工大學(xué);2001年

2 戴勇;數(shù)字半色調(diào)技術(shù)中誤差傳遞方法的研究[D];西安理工大學(xué);2001年

3 謝德紅;基于打印機(jī)和視覺模型的藍(lán)—綠噪聲抖動(dòng)加網(wǎng)算法的研究[D];武漢大學(xué);2005年

4 葉玉芬;數(shù)字半色調(diào)技術(shù)中的誤差擴(kuò)散算法研究[D];西安電子科技大學(xué);2006年

5 康莊;噴墨打印中墨滴落點(diǎn)的控制與補(bǔ)償[D];西安電子科技大學(xué);2009年

本文編號(hào)：2357730