A linguistic steganography based on word indexing compression and candidate selection

作者：Xiang, Lingyun^*; Wu, Wenshuai; Li, Xu; Yang, Chunfang

来源：Multimedia Tools and Applications, 2018, 77(21): 28969-28989.

DOI：10.1007/s11042-018-6072-8

• Linguistic steganography
• Steganalysis
• Compression
• Selection
• Synonym substitution
• Word indexing

In this paper, a novel linguistic steganography with high imperceptibility and undetectability is proposed via secret message compression and candidate text selection. The length of the practical embedded payload can be reduced by the proposed word indexing compression algorithm(WIC), while a best stego text with high undetectability can be selected from candidates by the stego text selection strategy. WIC algorithm losslessly compresses the secret message by combining a minimum maximum weight algorithm with Huffman coding under the help of the candidate cover text. To improve the anti-steganalysis capability, ten cover texts with small compression ratios are selected from a huge cover text set, and are embedded the corresponding compressed secret message by using synonym substitutions. Only one stego text is selected by a given rule derived from the distance between a cover text and its stego text. Experimental results show that the proposed compression algorithm achieves better compression ratios than Huffman and LZW coding algorithms leading to higher embedding efficiency, and our steganography performs well in anti-steganalysis capability with compression and the stego text selection rule.

• 出版日期2018-11
• 单位长沙理工大学