Detecting the LSB-LFSR Algorithm Stego-Image Using the Statistical Analysis and Visual Attack

Abstract

The aim of this paper is to detect the hidden information in BMP images when using Least Significant Bit (LSB) technique. In general, the passive steganalysis system deals with some stego-tools based on three stages:
Diagnosis, Breaking and Extraction. In this paper we focus in Diagnosis stage to attack a Linear Feedback Shift Register (LFSR) algorithm as stego tools. In the diagnosis process, the statistical analysis and visual test are suggested as tools for detection. These tests are: Mean Square Error (MSE) for stego-cover images, and Laplace Operator, chi-square analysis and visual test for stego only attack.