發(fā)布時(shí)間：2018-01-26 15:06

  本文關(guān)鍵詞： 嵌入式瀏覽器 渲染架構(gòu) 圖形引擎 硬件加速 CORDIC算法　出處：《電子科技大學(xué)》2014年碩士論文　論文類型：學(xué)位論文

【摘要】：隨著互聯(lián)網(wǎng)的飛速發(fā)展,HTML5標(biāo)準(zhǔn)的提出并廣泛被各種瀏覽器所支持,傳統(tǒng)的基于二維圖形的網(wǎng)絡(luò)頁(yè)面已經(jīng)不能滿足用戶的視覺(jué)和交互需求。在這一趨勢(shì)下,通過(guò)HTML5引入并實(shí)現(xiàn)的WebGL很好的解決三維圖形的繪制需要插件支持和難以跨平臺(tái)的問(wèn)題。嵌入式設(shè)備的飛速發(fā)展,例如平板、手機(jī)和智能電視的廣泛普及,使得用戶渴望時(shí)刻享受到來(lái)自互聯(lián)網(wǎng)應(yīng)用的便利。因此,基于呈現(xiàn)容器—瀏覽器的對(duì)于Web應(yīng)用程序的響應(yīng)速度和渲染效率直接影響著用戶的體驗(yàn)。但嵌入式設(shè)備有著匱乏的硬件資源,低功耗等特點(diǎn)。特別是缺乏3D圖形處理器或其它相關(guān)的協(xié)處理器,對(duì)于瀏覽器的圖形處理的優(yōu)化就顯得特別的必要。本文通過(guò)對(duì)現(xiàn)有的瀏覽器內(nèi)核WebKit的分析,研究基于WebKit的二維圖形與三維圖形渲染流程。實(shí)現(xiàn)一種與平臺(tái)和瀏覽器內(nèi)核無(wú)關(guān)的圖形庫(kù)架構(gòu)。WebKit通過(guò)該圖形庫(kù)渲染架構(gòu)向不同圖形引擎分發(fā)圖形渲染請(qǐng)求。針對(duì)特定平臺(tái)測(cè)試不同圖形處理引擎繪制效率,利用不同圖形庫(kù)的各自優(yōu)勢(shì)指定瀏覽器通過(guò)該架構(gòu)調(diào)用到指定的圖形渲染函數(shù),有效的提高了瀏覽器的渲染效率。同時(shí)運(yùn)用多線程技術(shù)在特定平臺(tái)上實(shí)現(xiàn)特定圖形庫(kù)的硬件加速。利用Direct FB和Skia混合渲染二維圖形的方式,更加合理的利用嵌入式設(shè)備的硬件資源,優(yōu)化用戶的體驗(yàn)。另外,針對(duì)部分嵌入式平臺(tái)缺乏浮點(diǎn)乘法處理器,而三維圖形渲染中涉及大量的三角函數(shù)的計(jì)算的特點(diǎn)。利用CORDIC算法僅使用加減和移位運(yùn)算的優(yōu)勢(shì),實(shí)現(xiàn)正余弦函數(shù)的計(jì)算,替換需要通過(guò)級(jí)數(shù)展開(kāi)計(jì)算的三角函數(shù),優(yōu)化在WebGL中的正余弦函數(shù),避免在計(jì)算三角函數(shù)的時(shí)候使用浮點(diǎn)運(yùn)算,降低計(jì)算復(fù)雜度,節(jié)約硬件資源,加速三維圖形在瀏覽器中的繪制。并對(duì)比測(cè)試該改進(jìn)算法的可行性。本文通過(guò)對(duì)當(dāng)前熱門(mén)的瀏覽器內(nèi)核與圖形庫(kù)的研究分析,提出的基于平臺(tái)無(wú)關(guān)的瀏覽器圖形庫(kù)架構(gòu),是利用多圖形引擎實(shí)現(xiàn)圖形渲染的一種嘗試。對(duì)于瀏覽器或其它需要圖形渲染的平臺(tái)進(jìn)一步設(shè)計(jì)優(yōu)化具有參考和借鑒意義。同時(shí),針對(duì)嵌入式平臺(tái)特定算法的優(yōu)化,為較低端的嵌入式設(shè)備實(shí)現(xiàn)三維圖形渲染的優(yōu)化,提供了一定的思路。
[Abstract]:With the rapid development of the Internet, the HTML5 standard is proposed and widely supported by all browsers, the traditional two-dimensional graphics, web page has been unable to meet the user's visual and interactive demand. Based on this trend, through the introduction and implementation of the HTML5 WebGL is a good solution to 3D graphics rendering plug-in support and difficult to cross platform the problem. The rapid development of embedded devices such as tablet, popularity of the mobile phone and smart TV, allows users to enjoy the Internet from time for the convenience of application. Therefore, the present container browser for Web applications based on the response speed and the rendering efficiency directly affects the user experience. But there is lack of embedded devices the hardware resources and low power consumption. Especially the lack of 3D graphics processor or other related coprocessor for optimizing graphics browser It is especially necessary. Based on the analysis of the existing WebKit browser kernel, WebKit of 2D graphics and 3D graphics rendering process. Based on the implementation of a graphics library architecture has nothing to do with the platform and browser kernel.WebKit through the graphics rendering architecture to different graphics engine distributed graphics rendering for a specific platform to test different requests. Graphics rendering engine, use of the advantages of different graphics library browser specified through the architecture call to the specified graphics rendering function, effectively improves the rendering efficiency. At the same time, the browser uses multithreading technology to achieve specific graphics library on a specific platform of hardware acceleration. The use of Direct FB and Skia hybrid two-dimensional graphics rendering method the embedded system hardware, more rational use of resources, optimize the user experience. In addition, some lack of floating platform Point multiplication processor, and the calculation of trigonometric functions involved in 3D graphics rendering. Taking advantage of CORDIC method using only addition and subtraction and shift operation, realize the calculation of sine and cosine function is replaced by series expansion, need trigonometric function calculation, optimization of sine and cosine functions in the WebGL, to avoid the use of floating-point arithmetic in computing trigonometric function, reduce the computational complexity, saving hardware resources, the acceleration of 3D graphics rendering in the browser. The feasibility and contrast test of the improved algorithm. Through the analysis of the current popular browser kernel and graphics library, platform independent graphics library architecture based on browser, is to use multiple graphics engine implementation try graphics rendering. For the browser or other graphics rendering platform for further design optimization has reference and significance. At the same time, according to the The optimization of the specific algorithm of the embedded platform provides a certain way of thinking for the optimization of the 3D graphics rendering of the lower end embedded devices.

【學(xué)位授予單位】：電子科技大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：TP393.092;TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前3條

1 伍鐵晟,盧延云;嵌入式Linux中圖形界面硬件加速的優(yōu)化設(shè)計(jì)[J];計(jì)算機(jī)工程與應(yīng)用;2004年33期

2 唐成戩;雷航;郭文生;;嵌入式瀏覽器軟硬件混合渲染層的研究與設(shè)計(jì)[J];計(jì)算機(jī)應(yīng)用;2013年05期

3 劉華星;楊庚;;HTML5——下一代Web開(kāi)發(fā)標(biāo)準(zhǔn)研究[J];計(jì)算機(jī)技術(shù)與發(fā)展;2011年08期

相關(guān)碩士學(xué)位論文 前1條

1 姜寧;嵌入式硬件圖形加速器的研究與設(shè)計(jì)[D];華東師范大學(xué);2006年

，

本文編號(hào)：1465913