ISE: AN INTELLIGENT AND EFFICIENT STEGANALYSIS ENGINE FOR IMAGE DATABASE IN BIG DATA SYSTEMS

Abstract

The aim of this work is to design a faster and artificially intelligent steganalysis engine, which is able to secure the image databases from any infected image in big data environment. The proposed Intelligent Steganalysis Engine (ISE) for image database in big data makes use of three steps, which are image estimation, feature generation and classification. In the first step, five new images are estimated from the original image, for computing 438 features and then these data images are passed through a classifier for final prediction of a stego image. The engine is designed based on Map-Reduce programming approach to cope with big data. The actual experiments were performed on the Big Data Hadoop by taking standard image data set. In the first two steps, the images are processed in both spatial and DCT domain. During these steps the implementations of image estimation and feature extraction algorithms become very much computationally intensive and seek a huge amount of time. The results obtained are compared with previously reported six similar works and an inference has been drawn for appropriate use of feature set and classifier pair.