透過您的圖書館登入
IP:3.145.152.242
  • 學位論文

運用影像辨識於視覺密碼應用系統之設計

Using Image Recognition Technology in Visual Cryptography Applications

指導教授 : 侯永昌

摘要


視覺密碼(Visual Cryptography)是一種將機密影像透過像素(pixel)層級的運算,加密成數張分享影像(Shares)的方法,而解密時,只要疊合足夠數量的分享影像,便能透過人類的雙眼,解開機密影像。以往分享影像是列印在透明膠片上,但有保存和攜帶不便,以及疊合時難以精確對準的問題。雖然到了網路時代,可以透過電腦解決疊合問題,但其它操作不便的情況依舊存在,導致視覺密碼學在實務應用上總是欠缺臨門一腳。 為了解決這樣的困境,本研究提出一項可行的應用系統架構,再透過影像辨識技術結合視覺密碼學,來開發一個具備創新特質的手機App,讓人手一支的智慧型手機,搖身一變成為「掃描即視」的解碼工具。有了這樣的工具後,這些視覺機密分享影像,就不用印在透明膠片上,或儲存在電腦裡,而是可列印在常見的普通文件紙、大型海報、或是直接顯示於數位媒體螢幕上。對於此點的改變,本研究更進一步提出新型的彩色機密分享影像製作方法,讓各種全彩影像都能化身為視覺密碼之祕密分享影像,呈現了不同以往的新面貌。這樣子一來,當想要查看隱藏的機密資訊時,就不用辛苦地去對準疊合透明膠片,只要拿起隨身的手機,往紙上、海報上或螢幕上輕輕一掃,機密影像立即浮現。 因此,本研究提出了一套可用在多數智慧型手機上的影像辨識步驟,並依此實作出一個具備「掃描即視」能力的Android App,可以顯示出隱藏在視覺機密分享影像裡的機密資訊。另外本研究也讓視覺密碼學不再侷限於資訊安全領域,將其應用於商品銷售推廣活動上,使之於成為電子折價券,並為多數人接納採用。

並列摘要


Visual cryptography is a type of cryptographic scheme, which hides secrets within multiple share images based at the pixel level. When a sufficient number of shares are stacked together, the secret can be recovered by the human eyes. These shares were conventionally printed out onto transparencies, but they encountered the inconvenience of keeping and carrying. The alignment difficulties also arose when superimposing transparencies. In the Internet era, although the problem of superimposing transparencies can be solved through computers, there are still other inconvenience interactions to suffer. This causes visual cryptography applications in practice always lacking the last push. To overcome the above obstacle, this study presents a feasible application schema for visual cryptography and develops a mobile app with innovations through image recognition technology. This allows each smartphone can be transformed into a decoding tool, which having the feature of “visible after scanning”. With this decoding tool, the visual secret sharing images do not need to be printed onto transparencies or continue saving in computers. Then they can be printed on plain paper, posters or billboards, or directly shown on the digital multimedia displays. For this change, this study further presents a novel method to generate the secret sharing image in full-color, which allows any full color pictures to be served as the cover images with significant visual meaning. It’s a new look different from the past, so you don’t need to superimpose transparencies in alignment difficulties anymore. Just scan the share images shown on paper or screen by picking up your smartphone whenever you want to recover the hidden secret. In this paper, we propose a mechanism for image processing which can be used on most smartphones, and also create an Android app having the feature of “visible after scanning”. This app can easily reveal the secret hidden within the share images. In addition, we provide a potential application for visual cryptography, so it is no longer limited to information security. It can be used as electronic coupons in marketing campaigns and accepted by most people.

參考文獻


4. 林金定,嚴嘉楓 & 陳美花 (2005)。質性研究方法:訪談模式與實施步驟分析。身心障礙研究季刊, 3(2), 122-136。
5. 林佳穎 (2012)。以機率分配為基礎的有意義視覺祕密分享。淡江大學資訊管理學系碩士班碩士論文。
1. Arias-Castro, E., & Donoho, D. L. (2009). Does median filtering truly preserve edges better than linear filtering?. The Annals of Statistics, 1172-1206.
2. Banville, C., & Landry, M. (1989). Can the Field of MIS be Disciplined?. Communications of the ACM, 32(1), 48-60.
4. Canny, J. (1986). A computational approach to edge detection. Pattern Analysis and Machine Intelligence, IEEE Transactions on, (6), 679-698.

延伸閱讀