圖像化QR code之灰階濃度調整於NFT郵票加值應用

Grayscale Adjustment Processing in Image-based QR Code for Value-Added Applications in NFT Stamps

隨著科技與日俱進，郵票種類也變得更為多元，除了傳統的紙製郵票，近年來也出現數位加密之NFT郵票。觀察各國郵政官方推廣的數位加密郵票，票上都會付有QR Code (Quick Response Code)，作為與數位網站的連結，因此將QR Code結合圖像，應用於NFT販售平台備受關注。然而，圖像化QR Code在輸出上容易受設備影響，導致其圖像不清晰，使肉眼辨識及機器讀取條碼資訊變得困難。因此本研究將通過不同灰階濃度對比，找出符合圖像化QR Code的最佳輸出之灰階濃度，同時因應不同類型輸出設備，做出相應調整。每次會將同灰階濃度的30張之圖像化QR Code以高品質與一般品質設備輸出，在進行模組錯誤率與碼字錯誤率分析，通過連續分析不同組之濃度以30張郵票為一組平均的辨認錯誤，來比較其是否在QR Code預設容錯範圍之內，藉此分析出機器可讀取且視覺呈現較佳的灰階濃度。經實驗與分析後發現灰階濃度變化會影響機器辨識率，因此根據圖像去調整不同的灰階濃度，即可達到圖像化QR Code最佳解碼率與良好的視覺樣貌，且研究之圖像化QR Code能實際整合於NFT郵票之加值產品。

As technology advances with each passing day, the variety of stamps has become more diverse. In addition to traditional paper stamps, digitally encrypted NFT stamps have also appeared in recent years. Observing the digital encrypted stamps officially promoted by postal offices in various countries, the stamps will be paid with a QR Code (Quick Response Code) as a link to the digital website. Therefore, the use of QR Code combined with images and applied to NFT sales platforms has attracted much attention. However, the output of graphical QR Code is easily affected by equipment, resulting in unclear images, making it challenging to recognize the barcode information with the naked eye and by machines. Therefore, this study will compare different gray-scale densities to determine the best output grayscale density that aligns with the graphical QR Code. Corresponding make corresponding adjustments will be implemented based on the different types of output devices. Each time, 30 graphical QR Codes with the same grayscale density will be output with high-quality and general-quality equipment, and the module error rate and code word error rate will then be analyzed. By continuously analyzing the density of different groups, 30 stamps will be used as the basis. A set of average recognition errors is used to compare whether they are within the preset error tolerance range of the QR Code, thereby analyzing the grayscale density that is machine readable and has better visual presentation. After conducting experiments and analysis, it was found that changes in gray-scale concentration will affect the machine recognition rate. Therefore, by adjusting different gray-scale concentrations according to the image, we can achieve the best decoding rate and improve the visual appearance of the graphical QR Code can be achieved, and the research is graphical. QR Code can actually be actually integrated into value-added products of NFT stamps.