A High Capacity and Imperceptible Text Steganography Using Binary Digit Mapping on ASCII Characters

摘要

Due to its light and multiplatform feature, plain text is widely used to transmit the information in news and social media online. However, it is vulnerable to attacks (e.g. unauthorized access or misuse of the information where the text is modified for particular purposes). To deal with the problem, secret text (ST) is embedded in overall plain texts appearing as a cover text (CT) so that any changes on CT can be detected. The present study proposes ST embedding method in CT by mapping ST binary digit onto binary digit of CT using ASCII characters-involving spaces, punctuation, and symbols. Prior to embedding process, the ST text was firstly encrypted with a One Time Pad (OTP) into ciphertext and each character was converted into binary number representing 7-bits long. Unlike the ST text, the CT text was immediately converted into 7 bits of binary number. The embedding process was conducted by mapping one bit of ST onto the first bit of CT character containing the same number of bit (e.g. mapping bit 1 of ST character onto the first bit 1 of CT character). Such a process was repeatedly carried out to ensure all bit of ST was completely embedded in CT. The system recorded each bit position as a stego key for any bit of ST had its position on bit of CT. The stego key served as a key to extract ST embedded in CT. The result of embedding process did not change the appearance of CT and therefore the method worked well with hidden information serving as text steganography or as watermark. In short, all CT characters can be used as a medium to hide ST where 1 character of ST required 7 characters of CT. In addition, the stego texts produced appeared identical to CT, as measured in similarity distance with Jaro-Winkler Distance of 1.