A Multi-Path Based Embedding Scheme at Perfect Maze

Abstract

Steganography is a process of hiding data in a media, e.g., image, audio, video, etc. There are lots of areas where steganography can be used. Many researchers are devoting their valuable efforts to hide messages before sending them to a destination through maze-based steganography. However, maze-based steganography has the drawback of reducing embedding capacity. Another problem of maze application at embedding is the imperfection of its solution paths. Imperfect paths are not used for embedding by the maze. That is why while selecting multiple paths, the process of Niwayama et al. in 2010 prunes many paths considering their overlapping or crossing affairs with other ones. In this paper, a solution to those stated problems of Niwayama et al. is outlined and successful results found are demonstrated here too. First, an imperfect maze is tried to be changed virtually to a perfect maze. Then solution paths from start to end are generated. While generating solution paths, the longest path among them is considered to cross or join a point along their pathways. Therefore, one path among those imperfect paths is considered embedded at and extraction was not employed by any earlier proposal, so far we know. Contribution to the stated research area will increase the embedding capacity. For measuring the longest path, the tie is broken by comparing x coordinates and then, if required, y coordinates. The illustrated result in the result analysis section details the justification of our claim.