Detecting Original Image Using Histogram, DFT and SVM

Information hiding for covert communication is rapidly gaining momentum. With sophisticated techniques being developed in steganography, steganalysis needs to be universal. In this paper we propose Universal Steganalysis using Histogram, Discrete Fourier Transform and SVM (SHDFT). The stego image has irregular statistical characteristics as compare to cover image. Using Histogram and DFT, the statistical features are generated to train One-Class SVM to discriminate the cover and stego image. SHDFT algorithm is found to be efficient and fast since the number of statistical features is less compared to the existing algorithm