Steganalysis Method for LSB Replacement Based on Local Gradient of Image Histogram

In this paper we present a new accurate steganalysis method for the LSBreplacement steganography. The suggested method is based on the changes that occur in thehistogram of an image after the embedding of data. Every pair of neighboring bins of ahistogram are either inter-related or unrelated depending on whether embedding of a bit ofdata in the image could affect both bins or not. We show that the overall behavior of allinter-related bins, when compared with that of the unrelated ones, could give an accuratemeasure for the amount of the embedded data. Both analytical analysis and simulationresults show the accuracy of the proposed method. The suggested method has beenimplemented and tested for over 2000 samples and compared with the RS Steganalysismethod. Mean and variance of error were 0.0025 and 0.0037 for the suggested methodwhere these quantities were 0.0070 and 0.0182 for the RS Steganalysis. Using 4800samples, we showed that the performance of the suggested method is comparable withthose of the RS steganalysis for JPEG filtered images. The new approach is applicable forthe detection of both random and sequential LSB embedding.