Title: An optimised approach to detect the identity of hidden information in grey scale and colour images
Authors: Murugeswari Kandavel; Deisy Chelliah; Ganesan Govindan
Addresses: Department of Computer Applications, Thiagarajar College of Engineering, Madurai, Tamil Nadu 625015, India ' Department of Computer Science and Engineering, Thiagarajar College of Engineering, Madurai, Tamil Nadu 625015, India ' Department of Computer Science and Engineering, K.L.N. College of Information Technology, Sivagangai, Tamil Nadu 630612, India
Abstract: Feature-based steganalysis is an emerging trend in the domain of Information Forensics, aims to discover the identity of secret information present in the covert communication by analysing the statistical features of cover/stego image. Due to massive volumes of auditing data as well as complex and dynamic behaviours of steganogram features, optimising those features is an important open problem. This paper focused on optimising the number of features using the proposed quick artificial bee colony (qABC) algorithm. Here we tested for three steganalysers, namely subtractive pixel adjacency matrix (SPAM), phase aware projection model (PHARM) and colour filter array (CFA) for the break our steganographic system (BOSS) 1.01 datasets. The significant improvement in the convergence nature of qABC quickly improves the solution and fine tune the search than their real counterparts. The results reveal that qABC method with support vector machine (SVM) classifier outperforms the non-optimised version concerning classification accuracy and reduced number of feature sets.
Keywords: steganalysis; feature selection; optimisation; classification.
International Journal of Business Intelligence and Data Mining, 2019 Vol.15 No.1, pp.1 - 21
Received: 12 Jan 2017
Accepted: 05 Apr 2017
Published online: 24 Apr 2019 *