1295134 九、發明說明： 【發明所屬之技術領域】 本創作係關於一種編碼及解碼的技術，特別是指一種以qr 為美 礎，提供一影像式編碼及解碼之方法。 【先前技術】 習用電子條碼係廣泛採用一維條碼之技術，國際常見之編碼方弋有 EAN-8、ΕΑΝ-13、UPC-A及UPC-E等各種標準，應用已相當廣泛；然而， 其内含之資訊量相當少，且只限於使用數字，在這資訊爆炸的時代已不敷 使用，由於資訊技術及硬體發展的進步，使我們得以擴展電子條碼的容量， 因而產生了二維條碼的標準及技術。常見的國際標準二維條碼有、 DataMatrix、MaxiCode及QRC〇de等各種標準，除了數字及符號外，有些 編碼方式還可以加入各國語言文字集，其中以QR c〇de最為大家所廣泛使 用，加上目前硬體設備發展突飛猛進，已經有快速且有效的二維條碼辨識 方案，二維條碼必定是將來重要的電子條碼技術。 習用之QRCode之編碼及解碼方法，主要包含QRC〇de編碼模組(13) 及QR Code解碼模組(14)所構成（如圖一所示）。qR c〇de編碼模組(13)接 受-串文字字串(11)，可以是數字、英文字、各種符號及各國語言文字的組 合，然後將之編碼成二進位值並轉換成影像圖案(12)。QR c〇de解碼模組則 接受一 QR Code影像(12)，利用影像處理的技術，將qR c〇de擷取出來並 將之解碼。 QR Code編碼模組(13)接受一串文字字串⑴"首先將文字分段並分析 (131) ’進行最佳化編碼方式（132)及錯誤位元修正碼（133) (ECC， 1295134 C—QKie)編碼方法。接著按照以上兩個編碼步驟後的二進位值字串 長度’來決定QRCode的影像大小(134)，最後在影像上依序填入二進位值 (135)，再加上遮罩(136)便產生QRC〇de影像(12)。 QRC〇de解碼模組⑼接受一 QRC〇de影像⑼，首先該影像先做二元 化處理(141)，使影像只有〇與丨兩個值。接著再試著定位蚊位圖案⑽） (FinderPattern ’屬於功能圖案的一種），並旋轉校正該影像(叫。再來谓 測是否有鮮酸(AlignmentPattern)來蚁獅(MGdd)(144)，並侧節奏 圖案(Timing PattemXUS)來切割QR Code影像(146)，再讀出遮罩資訊解除 遮罩(147)，最後依序讀出二進位值字串並解碼_，便可以得到原始字串 資訊(11)。 如果是使用資訊無須加密的編碼方式，以上標準的QRCOde編碼及解 碼就可適用，·然而，如將QRCGde應用在需要加密的資訊上時，則編碼及 解碼模組須進行更換調整，例如編碼模組在錯誤位元修正碼使用自訂的編 碼方式（如漢明碼，Hamming Code)，或是在輸入字串進行編碼前先行加密 (Encrypt)等方式，解碼模組亦進行相對應的調整，QR c〇de就可以有更 加廣泛的應用及可行性。 本案發明人鑑於QRCode具有高容量及廣泛流行的特性，乃亟思加以 改良創新，並經過多年苦心研究後，終於研發完成本件QR c〇de之編碼及 解碼方法，希望能將QRCode應用在電子票券及身份認證等各式各樣的產 業並加以推廣。 【發明内容】 本創作之目的即在於提供一 QR Code之編碼及解碼方法，提出方便的 1295134 、*方式’以及採用快速的影像處理技術解碼方法，以提出一有效、快速 之一維條碼演算法，並使QR Code廣泛應用於各個領域。 為達上述目的，本發明將標準之QR Code編石馬及解石馬模組更改如下（如 圖二所示）： 編碼模組含一加密（Encrypt)程序(231)，可採用各種可能之加密演算 法’以保護資訊的安全。 解碼模組含一解密（Decrypt)程序(249)，採用與加密演算法對應之解 密演算法。 本項發明可應用於電子票券或身份認證上，以電子票券為例，當饲服 器端接收到消費者索取電子票券之請求時（例如上網訂購演唱會門票），伺 服器便利用編碼演算法產生二維條碼（例如QRCode數位影像），然後再將 該二維條碼影像傳送至消費者的顯示設備上（可將QR Code影像傳送或儲 存於手機、PDA等可顯示QR Code之設備，以做為紙張門票的替代品）， 可視為一種電子票券；當消費者憑電子票券到具有二維條碼解碼演算法的 設備上，便可利用該設備讀出票券影像中所含有之訊息（例如使用QRCode 解碼設備辨識出該影像），不但節省消費者的排隊訂票、取票時間，並可以 兼顧環保，以減少資源浪費。 【實施方式】 請參閱圖二，本發明所提供之QRC〇de之編碼及解碼方法，包含下列 程序；首先’ QR Code編碼模組(23)接受一串文字字串(21)，先將文字字串 (21)進行三重資料加密標準(3-DES)或先進加密標準(AES)等方法加密 1295134 (231)，再敎字分段並分析(μ2);分段的方法可以是峡方式或雜方式， 固定方式指的是將文字分成數侧定長度的字串，轉性方式指的是將文 字分成長度不等的數辨φ ;分段完後再將各段字串進行分析，主要可將 字串分為四類：數字（Numeric)、英數字及符號（Alpha_Numeric)、八位元 文字（8_bitcharacter)及各國語言文字（BIG5、GB2312或8議瓜等各種 超過八位元的文字集），然後再將此四類進行最佳化的編碼方式Ο%)，此步 驟之目的係為了以最少的位元數來容納該資訊；編碼完畢後，接著再進行 錯誤位元修正碼(234) (ECC，E酣Correction c〇de)的編碼，國際標準使 用Reed Solomon演算法，此步驟之目的在於影像方式解碼時難免會有某些 位兀的二進位值讀取錯誤，因此需要錯誤位元修正碼來回復正確的二進位 值；接著按照以上兩個編碼步驟後的二進位值字串長度，來決定qr 的影像大小(235)，也就是所謂的Version大小，標準的QR c〇de有兩種模 式··模式1 (Model 1)及模式2(Model 2)，欲使用何種模式由系統使用者決定，1295134 IX. Description of the invention: [Technical field to which the invention pertains] This creation relates to a technique for encoding and decoding, and more particularly to a method for providing image encoding and decoding based on qr. [Prior Art] The conventional electronic bar code system widely adopts the technology of one-dimensional bar code. The international common coding methods include EAN-8, ΕΑΝ-13, UPC-A and UPC-E, and the application has been quite extensive; however, The amount of information contained in it is quite small, and it is limited to the use of numbers. It has not been enough in the era of information explosion. Thanks to advances in information technology and hardware development, we have been able to expand the capacity of electronic barcodes, thus generating 2D barcodes. Standards and techniques. Common international standard 2D barcodes, DataMatrix, MaxiCode and QRC〇de and other standards, in addition to numbers and symbols, some encoding methods can also be added to the national language text collection, which is widely used by QR c〇de, plus At present, the development of hardware devices is advancing by leaps and bounds. There are already fast and effective 2D barcode identification schemes. 2D barcodes must be an important electronic barcode technology in the future. The encoding and decoding method of the QRCode is mainly composed of a QRC〇de encoding module (13) and a QR Code decoding module (14) (as shown in Fig. 1). The qR c〇de encoding module (13) accepts a string literal string (11), which can be a combination of numbers, English words, various symbols, and various languages, and then encodes them into binary values and converts them into image patterns ( 12). The QR c〇de decoding module accepts a QR Code image (12) and uses the image processing technique to extract and decode the qR c〇de撷. The QR Code Encoding Module (13) accepts a string of text strings (1) " first segment and analyze the text (131) 'Optimize the encoding method (132) and error bit correction code (133) (ECC, 1295134 C - QKie) coding method. Then, according to the binary value string length ' after the above two encoding steps, the image size of the QRCode is determined (134), and finally the binary value (135) is sequentially filled in the image, and the mask (136) is added. Generate a QRC〇de image (12). The QRC〇de decoding module (9) accepts a QRC〇de image (9). First, the image is first binarized (141) so that the image has only two values, 〇 and 。. Then try to locate the mosquito pattern (10)) (FinderPattern 'is one of the functional patterns), and rotate to correct the image (call. Then check if there is AlignmentPattern to ants (MGdd) (144), and side The rhythm pattern (Timing PattemXUS) is used to cut the QR Code image (146), and then the mask information is unmasked (147). Finally, the binary value string is sequentially read and decoded _, and the original string information is obtained. 11) If the information is encoded without encryption, the above standard QRCOde encoding and decoding can be applied. However, if QRCGde is applied to the information that needs to be encrypted, the encoding and decoding module must be replaced and adjusted. For example, the encoding module uses a custom encoding method (such as Hamming Code) in the error bit correction code, or Encrypt before the input string is encoded, and the decoding module also corresponds to the decoding module. Adjustment, QR c〇de can have a wider range of applications and feasibility. In view of the high capacity and widespread popularity of QRCode, the inventor of this case has improved and innovated and has suffered for many years. After the research, I finally developed and coded the QR c〇de encoding and decoding method, hoping to apply QRCode to various industries such as e-tickets and identity authentication and promote it. [Invention] The purpose of this creation is Provide a QR Code encoding and decoding method, propose a convenient 1295134, *method' and use fast image processing technology decoding method to propose an effective and fast one-dimensional bar code algorithm, and make QR Code widely used in various fields. In order to achieve the above purpose, the present invention changes the standard QR Code coded stone and stone module to the following (as shown in FIG. 2): The coding module includes an encryption (Encrypt) program (231), which can adopt various possibilities. The encryption algorithm 'protects the security of the information. The decoding module includes a Decrypt program (249), which uses a decryption algorithm corresponding to the encryption algorithm. The invention can be applied to an electronic ticket or identity authentication. Taking an electronic ticket as an example, when the feeding device receives a request for a consumer to obtain an electronic ticket (for example, ordering a concert ticket online), the server facilitates encoding. The algorithm generates a 2D barcode (such as QRCode digital image), and then transmits the 2D barcode image to the consumer's display device (the QR Code image can be transmitted or stored in a device such as a mobile phone, PDA, etc. that can display the QR Code. As an alternative to paper tickets, it can be regarded as an electronic ticket; when the consumer uses an electronic ticket to a device with a two-dimensional barcode decoding algorithm, the device can be used to read the image contained in the ticket image. The message (for example, using the QRCode decoding device to recognize the image) not only saves consumers the time to queue tickets, collect tickets, but also balances environmental protection to reduce waste of resources. [Embodiment] Please refer to FIG. 2, the QRC〇de encoding and decoding method provided by the present invention includes the following procedures; first, the QR Code encoding module (23) accepts a string of text strings (21), first text The string (21) is encrypted by the Triple Data Encryption Standard (3-DES) or the Advanced Encryption Standard (AES), etc. 1295134 (231), and then segmented and analyzed (μ2); the segmentation method may be gorge mode or Miscellaneous mode, fixed mode refers to dividing the text into a number of sides of a certain length, and the translating method refers to dividing the characters into numbers of unequal lengths; after segmentation, the strings are analyzed. the string can be mainly divided into four categories: all kinds of more than eight yuan alphanumeric (numeric), alphanumeric characters and symbols (Alpha_Numeric), eight yuan text (8_bitcharacter) and national language (BIG5, GB2312, or 8 meeting melon Set), and then optimize the encoding of the four categories Ο%), the purpose of this step is to accommodate the information with a minimum number of bits; after the encoding is completed, then the error bit correction code ( 234) (ECC, E酣Correction c〇de) coding, international The Reed Solomon algorithm is used. The purpose of this step is that there will inevitably be some binary bit value reading errors when decoding in image mode. Therefore, the error bit correction code is needed to restore the correct binary value. Then follow the above two. The length of the binary value string after the encoding step determines the image size of the qr (235), which is the so-called Version size. The standard QR c〇de has two modes: Mode 1 and Mode 2 (Model 2), which mode to use is determined by the system user.
Model 1最多可容納4876個二進位值（Version 14)，而Model 2最多可容納 29648 個二進位值（Versions)。 在決定好QR Code影像大小之後，首先魅QR c〇de功能區參考圖 (Function Patterns Reference，如圖三所示），編碼時它用來區分功能圖案 (FunctkmPattems)及資料區域（EncodingRegi〇n)，解碼時可以方便定位 及取出資料編碼後的二進位值；由於資料係為二進位值，是以可使用一數 值來區分它為功能圖案或是資料區域，例如以1〇〇作為區分的數值，在功 月b區參考圖上某-點數值若為〇代表它是資料區域，若為謂則代表它是 1295134 功能圖案且二進位值為ο，若為101則代表它是功能圖案且二進位值為i ; 需要填入功能區參考圖之功能圖案有定位圖案（Finder pattems)、分隔線 (Separators)、節奏圖案(Timins Patterns)及調準圖案(AlignmentPatterns)或延 伸圖案(Extension Patterns)。 接著在>料區域内依標準的順序填入二進位值(236)，若填完後有剩下 的點未填入資料，則一律以〇值填入；最後再將QR c〇de影像(22)對八種 遮罩函數進行遮罩(237) (Masking)，然後在影像内記錄選用何種遮罩，此 步驟之主要目的係為使黑白點數量能儘量相等，並且避免一些黑白點的排 列方式。 QR Code解碼模組(24)接受一 QR Code影像(22)，首先該影像先做二元 化處理(241) ’使影像只有〇與1兩個值；在此使用改良過之κ中數叢聚法 (K-meanclustering)演算法作為二元化之方法，其中κ=2 ·· a.以影像的左上方點及中心點的灰階值（或彩色值）為初始的兩個群組 (group)中心值（Ck)。 b·影像上每一點所屬的群組(group)是以其灰階值mi (或彩色值）與兩 個群組(group)中心點的距離來決定，取歐基里德距離(Eudidean此加㈣較 小的群組(group)為所屬的群組(group);並計算新的中心值NCk = NCk + α (mi - NCk)。該新的中心值有初始值Nck = Ck。 c. 更新兩個群組(gr0Up)的中心值，並計算其差值dk ; Ck = NCk，业= 丨 NCk Ck 丨。 d. 重複b，c兩步驟，直到所有dk</5或重複次數達到7為止。 1295134 接著再4著疋位出疋位圖案(242) (Finder p姐咖，屬於功能圖案的一 種）找到一個定位圖案後，就可以決定該吸㈤e影像的旋轉角度，並旋 轉校正該賴243); _校正後賊__相雖一邮pattem)(專 屬於M〇del2的功㈣案），進行模福)及影像(如㈣麟(244)，若 是找到了貝h亥QR Code影像應為qr c〇de M〇dd 2影像，否則應為qR Code Model 1 Model 能圖案的-種）並計算其數量，可用來確認該QRQ)de影像的ϋ〇η大小 (245)，並用來切割該QRC〇de影像(246)，將切割的影像決定每一方格之二 進位值’項出遮罩資崎遮罩解除，再依標準的順序讀出三進位值(24乃， 最後再進行錯錄元修正簡碼(248)，崎行文字職(249)，便可得到原 本所輸入之文字字串(21)。 實施例一 1· QR Code編碼模組部份 明參閱圖四，輸入文字字串為“AC_42”（41)，經過3_DES文字加密、 文字分段及分析、資料編碼及加入錯誤位元修正碼四個步驟(421)後得到二 進位值字串 “〇〇1000000101〇〇111〇111〇〇111〇〇1〇〇〇〇1〇”（422)，此字串之長 度為35 ’我們選用M〇del 1為例子，評估字串長度之後選用Versk^ ^3)， 並產生功能區參考圖(424)，接著再以M〇del丨的順序將二進位值字串填入 (425);剩下未填入的那些點，一律以〇值填入(426)，然後再經過遮罩處理 (427)選擇最適當的遮罩影像(43)為最後輸出。 2· QR Code解碼模組部份 1295134 請參閱圖五，輸入一 QR C〇de影像(51)，經過二元化的步驟(521)後得 到一元化影像(522)。再按照定位圖案的特性（b:w:b:w:b=Li:3:i:i，b代表 二進位值為1的點數，w代表二進位值為〇的點數），找出三個定位圖案的 中心(523)(圖中灰色横線及縱線的交點524)，經過模式(M〇(jel)及影像 (Version)的判斷，可以得知此影像為M〇dd丨的QRC〇de (因為圖中沒有調 準圖案，Alignment Pattern)，並且為Version 1的大小，接著再找節奏圖案 (Timing Pattern)用來確認Version大小並用以切割影像(525)，然後從影像 (526)中讀遮罩資喊嫌鮮，最後讀&二進健並進行解碼及文字解 密(527)可得到該影像資訊(528)為“AC_42”（53)。 本創作所提供之QR Code之編碼及解碼方法，射述其他制技術相 比較時，更具有下列之優點： 1·可進行文字加密及解密技術，以確保資訊安全。 2.提供快速影像處理演算法，並且毋需列印二維條鄕像即可辨識， 可避免紙張浪費。 上述詳細說明係針對本發明之-可行實施例之具體說明，若依本發明 構想所作之改變，其產生之功能作用，仍未超出綱與圖示所涵蓋之精神 時，均應包含在本創作之範圍内，合予陳明。 综上所述，本案的雜卵技術增進上述多項功效，應已充分符合新 穎性及進频之法定發明·，爰錄提㈣請，騎貴局核淮本件_ 專利申請案，以勵發明，至感德便。 【圖式簡單說明】 1295134 圖一為習用QRCode編碼及解碼模組流程圖； 圖二為本發明之QRCode編碼及解碼模組流程圖； 圖三為以模式l(Modell)，影像2(Version2)為例的功能區參考圖； 圖四為本發明QR Code編碼實例； 圖五為本發明QR Code解碼實例。 【主要元件符號說明】 11文字字串 12 QR Code 影像 13 QR Code編碼程式 14 QR Code解碼程式 21文字字串 22 QR Code 影像 23 QR Code編碼程式 24 QR Code解碼程式 41文字字串 42 QR Code編碼模組 43遮罩影像 51遮罩影像 52 QR Code解碼模組 53文字字串Model 1 can hold up to 4876 binary values (Version 14), while Model 2 can hold up to 29648 binary values (Versions). After determining the QR Code image size, firstly, the QR C〇de function area reference map (shown in Figure 3) is used to distinguish between the functional pattern (FunctkmPattems) and the data area (EncodingRegi〇n). When decoding, it is convenient to locate and take out the binary value of the data encoding; since the data is a binary value, it can be distinguished by using a numerical value as a functional pattern or a data area, for example, a value determined by 1〇〇. In the reference picture of the power month b area, if the value of a point is 〇, it means that it is the data area. If it is said, it means that it is 1295134 and the binary value is ο. If it is 101, it means that it is a functional pattern and two. The carry value is i; the functional patterns that need to be filled in the functional area reference map are Finder pattems, Separators, Timins Patterns, and Alignment Patterns or Extension Patterns. Then, in the > material area, the binary value (236) is filled in the standard order. If the remaining points are not filled in the data after filling, the data will be filled in with the 〇 value; finally, the QR c〇de image will be added. (22) Mask (237) (Masking) the eight mask functions, and then record which mask is selected in the image. The main purpose of this step is to make the number of black and white points as equal as possible and avoid some black and white points. Arrangement. The QR Code decoding module (24) accepts a QR Code image (22). First, the image is first binarized (241) 'The image has only two values and one value; the modified κ median cluster is used here. The K-meanclustering algorithm is used as a method of binarization, where κ=2 ·· a. takes the gray-scale value (or color value) of the upper left point of the image and the center point as the initial two groups ( Group) Center value (Ck). b. The group to which each point on the image belongs is determined by the distance between the grayscale value mi (or color value) and the center point of the two groups. Take the Euclid distance (Eudidean this plus) (4) The smaller group is the group to which it belongs; and the new central value NCk = NCk + α (mi - NCk) is calculated. The new center value has the initial value Nck = Ck. The center value of two groups (gr0Up), and calculate the difference dk; Ck = NCk, industry = 丨NCk Ck 丨 d. Repeat b, c two steps until all dk < /5 or the number of repetitions reaches 7 1295134 Then, after the 疋 position pattern (242) (the Finder p sister, which belongs to the functional pattern), after finding a positioning pattern, the rotation angle of the suction (five) e image can be determined, and the rotation is corrected. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ For the qr c〇de M〇dd 2 image, otherwise it should be the qR Code Model 1 Model can be used to determine the number of the QRQ) de image ( 245), and used to cut the QRC〇de image (246), and the cut image determines the binary value of each square' item to remove the mask, and then read the triple value in the standard order (24) Finally, the wrong record correction short code (248) and the Japanese character (249) can be used to obtain the originally entered text string (21). Embodiment 1 1. QR Code Encoding Module Partially Referring to Figure 4, the input text string is "AC_42" (41), and the binary value string is obtained after four steps (421) of 3_DES text encryption, text segmentation and analysis, data encoding, and adding error bit correction code. 〇〇1000000101〇〇111〇111〇〇111〇〇1〇〇〇〇1〇”(422), the length of this string is 35 'We use M〇del 1 as an example. After evaluating the string length, use Versk^ ^3), and generate a functional area reference map (424), and then fill in the binary value string in the order of M〇del丨 (425); the remaining unfilled points are always filled with 〇 value (426), then the masking process (427) is followed to select the most appropriate mask image (43) as the final output. 2· QR Code Decoding Module Part 1295134 Please refer to Figure 5, input a QR C〇de image (51), and after the binarization step (521), obtain a unified image (522). According to the characteristics of the positioning pattern (b:w:b:w:b=Li:3:i:i, b represents the number of points with a binary value of 1, and w represents the number of points with a binary value of 〇), find out The center of the three positioning patterns (523) (the intersection of the gray horizontal line and the vertical line in the figure 524), after the mode (M〇 (jel) and the image (Version) judgment, it can be known that the image is M〇dd丨 QRC 〇de (because there is no alignment pattern, Alignment Pattern), and the size of Version 1, then find the Timing Pattern to confirm the Version size and cut the image (525), then from the image (526) The middle reading mask is screaming, and finally reading & binary and decoding and text decryption (527) can obtain the image information (528) as "AC_42" (53). QR Code encoding provided by this creation And the decoding method, when comparing other technologies, has the following advantages: 1. Text encryption and decryption technology can be used to ensure information security. 2. Provide fast image processing algorithm, and need to print 2D The image can be identified to avoid paper waste. The above detailed description is for the present invention - feasible The specific description of the example, if it is changed according to the concept of the present invention, the function of the invention should not be included in the scope of this creation, and should be included in the scope of this creation. As mentioned above, the technology of the eggs in this case enhances the above-mentioned multiple functions, and should have fully complied with the novelty and the legal invention of the frequency. · 爰录提(4) Please, ride your nuclear medicine _ patent application, in order to invent, to Figure 1 is a flow chart of the QRCode encoding and decoding module; Figure 2 is a flow chart of the QRCode encoding and decoding module of the present invention; Figure 3 is a model l (Modell), image 2 (Version 2) is an example of a functional area reference diagram; FIG. 4 is an example of QR Code encoding according to the present invention; FIG. 5 is an example of QR Code decoding according to the present invention. [Description of main component symbols] 11 text string 12 QR Code image 13 QR Code code Program 14 QR Code Decoding Program 21 Character String 22 QR Code Image 23 QR Code Encoding Program 24 QR Code Decoding Program 41 Text String 42 QR Code Encoding Module 43 Mask Image 51 Mask Image 52 QR Code Decoding Module 53 Character string