多秘密視覺密碼之研究

• Main Authors: 朱致諺, Chu, Chih Yen
• Language: 中文
• Published: 國立政治大學
• Subjects: 視覺密碼; 多秘密; 像素擴展
• Online Access: http://thesis.lib.nccu.edu.tw/cgi-bin/cdrfb3/gsweb.cgi?o=dstdcdr&i=sid=%22G0098753029%22.

視覺密碼不同於傳統密碼學，在於視覺密碼不需經過繁雜的計算，而是利用分享影像間的疊合，再透過人類視覺得到原先所要隱藏的秘密影像。傳統視覺密碼被用來設計在隱藏單張秘密影像，但對於隱藏多張秘密影像非常沒有效率，因此本文希望僅以兩張分享影像間的疊合，使用上下或左右平移的方式解密更多張疊合後的秘密影像。 多數的多秘密視覺密碼技術，隱藏的數量侷限在分享影像旋轉的次數，即使目前已有技術可隱藏到n張秘密影像，但對比及形狀比例皆會隨著隱藏的數量作改變。對比決定影像清晰度的關鍵因素；像素擴展決定疊合影像的大小，因此這兩項一直是視覺密碼重要的研究指標。本文主要貢獻在於：(1) 隱藏的數量不受侷限 (2) 對比、像素擴展及形狀比例不隨著隱藏數量改變(3) 產生最小面積的疊合影像。 === Visual cryptography is different from the conventional cryptography, because it can be decrypted directly by human visual system without any complex computation. Conventional visual secret sharing is designed for hiding only single secret image, and it is inefficient to hide numerous secret images. In this thesis, we introduce a new visual secret sharing scheme which requires only two share images, and by shifting one of them up to down or left to right, our new scheme can decrypt more secret images. In most of the current researches on multiple secrets visual cryptography, the amount of the images that can be hidden is limited by the number of times that a sharing image can be rotated. Although some researches provide ways to hide n secret images where n can be any value, the contrast and the ratio of shape will decrease when n is large. Because the contrast determines the image’s clarity and pixel expansion determines the size of composite image, they are two important indicators of researches on visual secret sharing. The main contributions of this thesis are: Firstly, the amount of secret images to be hidden can be any value n. Secondly, the contrast, pixel expansion and the ratio of shape is independent of n, which means that they will remain the same for different n. Thirdly, when hiding the same amount of secret images, our scheme has the least size of composite image (ie., the size of share B) comparing with other schemes.