本文選題：工業(yè)QR碼 + 自適應(yīng)伽馬校正　； 參考：《安徽大學(xué)》2017年碩士論文

【摘要】：21世紀(jì),是信息化的社會(huì),信息化社會(huì)帶給人們更方便快捷的生活方式。圖像作為人類(lèi)感知世界、了解世界的視覺(jué)根基,是人類(lèi)獲取信號(hào)、表達(dá)情感和傳遞溝通的重要手段,二維碼如QR,Data Matrix,Aztec Code,Maxicode等,作為符合以上所有特征的符號(hào),被研發(fā)之初主要應(yīng)用在汽車(chē)裝配期間汽車(chē)零件的跟蹤,之后的二十多年間,被擴(kuò)展應(yīng)用在各個(gè)行業(yè)中。除此之外,二維碼自身所具備的特點(diǎn),如:低成本、高可靠性、高保密性、高糾錯(cuò)能力、可編輯外觀等,也是其被廣泛應(yīng)用的主要原因。傳統(tǒng)流通過(guò)程中的QR碼多屬于半自動(dòng)的識(shí)別,如生活超市中的QR碼在第一次掃描失敗后,可以更正識(shí)別方向繼續(xù)識(shí)別,從而容易檢測(cè)出符號(hào)中蘊(yùn)含的信息。但工業(yè)生產(chǎn)線上的QR碼識(shí)別,不可輕易改變識(shí)別條件,所以識(shí)別難度相對(duì)較大。本文的主要研究對(duì)象是工業(yè)二維碼。文章介紹了 QR條碼圖像的基礎(chǔ)知識(shí)和碼制特征以及編譯碼原理,并對(duì)糾錯(cuò)碼在解碼過(guò)程中的步驟有了詳細(xì)了解。QR碼圖像識(shí)別包括二維碼符號(hào)鎖定,圖像清晰處理以及解碼等步驟完成,在本文中,研究了二維QR碼處理的基本過(guò)程,并研究了以下的內(nèi)容:(1)針對(duì)工業(yè)QR二維碼圖像采集環(huán)境受限,導(dǎo)致二維碼圖案質(zhì)量不佳,包括背景過(guò)暗或曝光過(guò)強(qiáng)等情況,采用自適應(yīng)伽馬校正方法,增加圖像對(duì)比度。(2)針對(duì)工業(yè)QR碼圖像背景復(fù)雜,干擾因素較多,使用雙立方插值等算法進(jìn)行無(wú)失真縮放。(3)針對(duì)工業(yè)QR二維碼符號(hào)壓縮性強(qiáng),在巨大的零件上占空比小的問(wèn)題,依據(jù)QR二維碼自身符號(hào)特征,采用圖像模板匹配搜索的方式進(jìn)行二維QR碼輔助搜索。在QR碼的識(shí)別過(guò)程中,傳統(tǒng)的圖像算法對(duì)平面圖像有很好的處理效果,當(dāng)這些算法應(yīng)用在二次曲面失真QR碼時(shí)識(shí)別率就不是理想中的那么好。為此,本文提出—種結(jié)合QR碼自身符號(hào)特征—位置探測(cè)圖案,基于曲線擬合的曲面QR碼校正算法,該算法能夠?qū)崿F(xiàn)對(duì)曲面QR碼快速精確校正的處理結(jié)果。本篇文章整合經(jīng)典圖像預(yù)處理的算法以及本文提出的校正算法,使用C++語(yǔ)言編程,在結(jié)合了 Google開(kāi)放源碼庫(kù)ZXing的基礎(chǔ)上,設(shè)計(jì)了針對(duì)多種異常環(huán)境下采集的圖像的軟件,經(jīng)過(guò)在本軟件以及其他軟件上數(shù)百?gòu)垐D片的實(shí)驗(yàn),本軟件的識(shí)別率以及識(shí)別時(shí)間相比較于其他識(shí)別軟件有明顯的提升。鑒于一般軟件無(wú)法識(shí)別曲面QR碼,所以對(duì)曲面QR碼單獨(dú)進(jìn)行實(shí)驗(yàn),驗(yàn)證當(dāng)圖像以不同尺寸處于不同曲率的曲面上時(shí),曲線擬合算法對(duì)圖像的可識(shí)別性。
[Abstract]:In the 21 st century, it is an information society, which brings people a more convenient way of life. As the visual basis of human perception and understanding of the world, image is an important means for human beings to acquire signals, express emotions and transmit communication. QR codes, such as QRX data Matrix Aztec CodeMaxicode, are symbols that conform to all of the above characteristics. It is mainly used in auto parts tracking during automobile assembly at the beginning of R & D, and has been extended to various industries for more than 20 years. In addition, the features of QR codes, such as low cost, high reliability, high confidentiality, high error-correcting ability, editable appearance and so on, are also the main reasons for its wide application. The QR codes in the traditional circulation process are mostly semi-automatic recognition. For example, the QR codes in the life supermarket can correct the recognition direction and continue to recognize after the first scan failure, so that the information contained in the symbol can be easily detected. However, QR code recognition in industrial production line can not easily change the recognition conditions, so it is relatively difficult to identify. The main research object of this paper is industrial two-dimensional code. This paper introduces the basic knowledge of QR barcode image, the character of code system and the principle of encoding and decoding, and also gives a detailed understanding of the steps of error-correcting code in the decoding process, including the two-dimension code symbol lock in the image recognition of .QR code. In this paper, the basic process of two-dimensional QR code processing is studied, and the following contents: 1) aiming at the limitation of the image acquisition environment of industrial QR QR code, the quality of QR code pattern is not good. The background is too dark or the exposure is too strong. The adaptive gamma correction method is used to increase the contrast of the image.) for the industrial QR code image, the background is complex, and the interference factors are many. In order to solve the problem of high compression and low duty cycle of industrial QR QR codes, based on the symbol characteristics of QR QR codes, we use bicubic interpolation and other algorithms to scale without distortion. The image template matching search method is used to carry out two-dimensional QR code aided search. In the process of QR code recognition, the traditional image algorithm has a good processing effect on plane image. When these algorithms are applied to Quadric distortion QR code, the recognition rate is not as good as ideal. Therefore, this paper presents a curved surface QR code correction algorithm based on curve fitting combined with QR code's symbol feature and position detection pattern. The algorithm can realize the fast and accurate correction of surface QR code. This paper integrates the classical image preprocessing algorithm and the correction algorithm proposed in this paper, using C language programming, combined with the Google open source code library ZXing, designed the image acquisition software for a variety of abnormal environments. Through the experiment of hundreds of pictures on this software and other software, the recognition rate and recognition time of this software are obviously improved compared with other recognition software. Because the general software can not recognize the surface QR code, the surface QR code is tested separately to verify the recognition ability of the curve fitting algorithm when the image is on a curved surface with different sizes and different curvature.
【學(xué)位授予單位】：安徽大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類(lèi)號(hào)】：TP391.4

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 吳俊斌;吳晟;吳興蛟;;矢量填充和插值算法的圖像放大[J];計(jì)算機(jī)與數(shù)字工程;2016年06期

2 鄭學(xué)敏;;條碼技術(shù)在物流管理中的應(yīng)用分析[J];物流科技;2016年03期

3 劉菁欣;白云;王俊;;改進(jìn)霍夫變換的樞紐立交橋檢測(cè)方法[J];測(cè)繪科學(xué);2016年10期

4 劉烽杰;蔡明;;柱面QR碼的識(shí)別及實(shí)現(xiàn)[J];計(jì)算機(jī)與現(xiàn)代化;2015年02期

5 林天圓;王杰;李金屏;;一種光照不均勻圖像的灰度校正方法[J];濟(jì)南大學(xué)學(xué)報(bào)(自然科學(xué)版);2015年06期

6 史志鋒;;基于三維透視變換的圓柱面QR碼識(shí)別方法[J];現(xiàn)代電子技術(shù);2014年08期

7 劉寧鐘;葉超;蘇軍;;基于不變矩特征的二維碼模糊類(lèi)型辨識(shí)算法[J];數(shù)據(jù)采集與處理;2014年01期

8 田軍;孟祥娟;王萍;;全景圖中投影模型與算法[J];計(jì)算機(jī)系統(tǒng)應(yīng)用;2013年05期

9 馮春貴;祝詩(shī)平;王海軍;賀園園;;基于改進(jìn)模板匹配的限速標(biāo)志識(shí)別方法研究[J];西南大學(xué)學(xué)報(bào)(自然科學(xué)版);2013年04期

10 謝榮生;黃鵬程;許高攀;;QR碼電子票務(wù)系統(tǒng)數(shù)字水印防偽方法設(shè)計(jì)[J];廈門(mén)理工學(xué)院學(xué)報(bào);2013年01期

，

本文編號(hào)：1931300