Median Filtering Detection Using Markov Process in Digital Images

Digital images are prevalent medium for information representation. The wide acceptability of images also leads to some problems. Few people spread forged images for their own interests. This type of practice questions the credibility of images. To ensure the credibility, forensic investigation is carried out to ensure the genuineness of images. Detection of median filtering is one of the important forensic analyses. Median filtering has several uses in antiforensics. It is applied to hide the impression of image forgery due to its nonlinear nature. Some techniques are evolved to countering these antiforensics i.e., median filtering. These techniques can classify median filter operated and non-median filter operated images. In this paper, we compare three techniques that are based on Markov process. We find that performance of these techniques varies according to the size of database.