紅外線 QR Code 技術應用於 NFT 之研究

Abstract

二維條碼常用於行動支付、商品連結及資訊儲存，最廣為人知的二維條碼為QR Code(Quick Response Code)，受到市場美學行銷及保護個人重要資訊的意識興起，QR Code外觀美化及安全防偽所帶來的附加價值逐漸受到重視。加密貨幣與非同質化貨幣（Non-fungible token，NFT）風靡全球，此類型線上貨幣皆需要使用加密錢包來儲存資產，若要與實體商品相互結合，大多是將加密錢包私鑰或助記詞列印下來或是製成二維條碼。因此本研究利用C（Cyan，青色）、M（Magenta，洋紅色）、Y（Yellow，黃色）在紅外線光源下會呈透明狀來製作外顯圖像化 QR Code，將外顯的QR Code與NFT圖像做結合，達到美化QR Code的效果。並將加密錢包中的助記詞以紅外線QR Code技術藏入NFT外顯QR Code當中，利用K(Black）中的碳黑油墨會吸收紅外光的特性，可以在紅外線光源下顯現影像之原理製作內藏之助記詞QR Code，以提升其防偽安全之特性。本研究針對字元數100之資訊量，選擇三種版本之QR Code，分別為第四版（33×33）容錯等級L（7%）、第六版（41×41）容錯等級Q（25%）以及第八版（49×49）容錯等級H（30%）。擷取紅外線下的影像後，使用Matlab程式將其影像強化，再利用資訊點識別程式進行錯誤率分析，找出適用於紅外線下之QR Code版本與容錯等級搭配方式。結果顯示使用Matlab影像強化能夠大幅度降低錯誤率，且相同尺寸下QR Code版本越小Module錯誤率越低。而本研究最小版本第四版（33×33）Codeword錯誤率僅8%，但其容錯等級為L（7%），難以被正確識別讀取內容，因此在實用上第六版較適合藏入紅外線圖像化QR Code。本研究提出之方法除了有效降低紅外線QR Code的Module錯誤率及Codeword錯誤率外，亦能找出適用藏入紅外線圖像化QR Code之版本及容錯等級，讓使用者在離線儲存錢包助記詞的同時，兼具防偽安全的特性及精準讀取的方便性。Two-dimensional barcodes are widely utilized for various applications, including mobile payments, product linking, and information storage, with the QR Code (Quick Response code) standing out as the most recognized. As the awareness of market aesthetics, marketing strategies, and the protection of personal information continues to grow, there is an increasing emphasis on enhancing the visual appeal and security features of QR Codes.The global surge in popularity of cryptocurrencies and non-fungible tokens (NFTs) has made encrypted wallets essential for secure asset storage. When these digital currencies integrate with physical goods, it is customary to print or generate visually appealing QR Codes containing the private key or mnemonic of the crypto wallet. Consequently, this study aims to capitalize on the transparency of Cyan (C), Magenta (M), and Yellow (Y) under infrared light to create visually enhanced QR Codes. These aesthetically pleasing QR Codes are then merged with NFT images, elevating the overall visual appeal. Additionally, the study involves concealing mnemonic QR Codes within the visual QR Codes using infrared technology. This takes advantage of the properties of Carbon Black ink in the Key (Black) component, which absorbs infrared light and reveals the embedded mnemonic. Such an approach enhances both the security and anti-counterfeiting features.For an information content of 100 characters, the study selects three QR Code versions: version 4 (33×33) with an error correction level of L (7%), version 6 (41×41) with an error correction level of Q (25%), and version 8 (49×49) with an error correction level of H (30%). Following the capture of infrared images, Matlab is employed to enhance them. Subsequently, error rate analysis is performed using information point recognition programs to determine the optimal QR Code version and error correction level for infrared applications. The results indicate that Matlab image enhancement significantly reduces error rates, with smaller Module sizes leading to lower error rates. However, the smallest version 4 (33×33) has an error rate of 8% with an error correction level of L (7%), making it challenging to be correctly identified. Therefore, for practical use, version 6 is deemed more suitable for embedding in infrared QR Code.The proposed method not only effectively reduces Module and Codeword error rates in infrared QR Codes but also identifies the suitable version and error correction level for embedding. This provides users with both anti-counterfeiting security and convenient offline storage of wallet mnemonics.