Robust First Quality Factor Estimation For Double Compressed and Resized Images

Abstract

The ability to identify tampering in JPEG images has become increasingly significant due to the high usage of JPEG as a compression technology. In doubly compressed and resized (DCR) images, this paper revisits the state-of-the-art resampling factor estimate and estimation of primary quality factor. As a result of the recent findings in the literature, the proposed technique improves existing approaches for analyzing double JPEG compression in the presence of resizing in between two compressions. By utilizing the period of Welch Power spectral density (PSD) of reverse resized image to estimate primary quality factor which is an important forensic clue of processing history of an image. Empirically find the nonlinear relationship between the first quality factor and location of the most prominent peak, which is the period of DC histogram of the first compression by utilizing Welch PSD analysis. Experimental validation of the proposed method is carried out on several JPEG images, and comparing with existing state-of-the-art methods reveals that the proposed method performs better.