A hybrid feature selection approach based on improved PSO and filter approaches for image steganalysis
- PDF / 835,214 Bytes
- 12 Pages / 595.276 x 790.866 pts Page_size
- 113 Downloads / 255 Views
ORIGINAL ARTICLE
A hybrid feature selection approach based on improved PSO and filter approaches for image steganalysis Rita Rana Chhikara1 • Prabha Sharma1 • Latika Singh1
Received: 8 June 2015 / Accepted: 14 October 2015 Springer-Verlag Berlin Heidelberg 2015
Abstract This paper proposes a novel feature selection approach to improve the classification accuracy and reduce the computational complexity in image steganalysis. It is a hybrid filter-wrapper approach based on improved Particle Swarm Optimization (PSO). It consists of two phases: the first phase is composed of two filter techniques namely t test and multiple-regression which selects the features based on their ability to discriminate images as stego or cover. The second phase further reduces the number of features by working on the significant features selected during the first phase using an improved PSO. This approach overcomes the disadvantages of global best PSO by integrating it with local best PSO and dynamically changing the population size (Hope/Rehope). The proposed approach is tested on two sets of features extracted from spatial domain (SPAMSubtractive Adjacency Matrix) and transform domain (CCPEV-Cartesian Calibrated features extracted by Pevny´) for four embedding algorithms nsF5, Outguess, Perturbed Quantization and Steghide using SVM (Support Vector Machine) classifier. Experimental results demonstrate that this approach significantly improves the
& Rita Rana Chhikara [email protected] Prabha Sharma [email protected] Latika Singh [email protected] 1
School of Engineering and Technology, ITM University, Gurgaon 122017, India
classification accuracy and drastically reduces dimensionality as compared to results produced by some wellknown feature selection algorithms. Keywords Image steganalysis Hybrid feature selection technique Particle swarm optimization t-test Multiple regression
1 Introduction In today’s digital world, where communication mostly happens through internet, steganography has become cause of concern for law enforcement agencies. It is the art and science of hiding secret messages by modifying the redundant bits in the cover medium such as image, audio, video or text files [1]. To track the misuse of this covert communication, Steganalysis, a science to detect presence of hidden information, has gained equal importance. Steganalysis has become a challenging task in recent years and researchers have attacked it as a two class pattern recognition problem, which involves extracting significant features from images and classifying them as ‘stego’ or ‘cover’ based on presence or absence of hidden message [36]. Different classification algorithms like Fisher Linear Discriminant, Support Vector Machine, and Neural Network have been employed in literature to classify images as cover or stego [6, 10, 31]. In order to achieve optimal classification accuracy high dimensional features from different domains have been constructed. For example, SPAM [38] in spatial domain, Markov features merged with
Data Loading...
