Vol. 2, Issue 5 (2017)
Feature selection for steganalysis using glow worm algorithm
Author(s): Neha Singh, Jyoti Kumari, Charu Aggarwal
Abstract: Steganography is the process of communicating the way in which a secret message is hidden in some other information. Steganalysis is the art of finding the hidden messages inside digital data, if exists. Feature based steganalysis is a branch in information forensics. Its fundamental point is to identify the nearness of an incognito correspondence by utilizing the factual features of the cover and stego image as pieces of information. Because of the substantial volumes of security review information and additionally mind boggling and dynamic properties of steganogram practices, enhancing the execution of steganalysers turns into an imperative open issue. To enhance the execution of steganalyser, the order procedure is streamlined by utilizing less number of features required for the arrangement. This is finished utilizing feature determination. In the proposed paper, steganalysis of pictures is done and four distinctive feature sets were utilized for the examination. ELM is utilized as the classifier for steganalysis and Glow-worm Swarm Optimization Algorithm was effectively actualized for includes determination. The test comes about demonstrate that there has been an extraordinary lessens the quantity of features being chosen and the classiﬁcation rate has moved forward.