Universal Method for Detecting Violations in the Integrity of a Digital Image Based on Analysis of Blocks of its Matrix

Abstract

The requirement to provide the content expertise (particularly digital video) in real time is becoming critical. Thus, the aim of the work is to increase the efficiency of identifying the fact of violation of image integrity by developing a universal expert method with low computational complexity. This aim was achieved by using a new approach developed by the authors earlier that based on the properties of the dependence of the frequency index of the singular vector of the image matrix on its number, adapted for the case of a block organization of expertise. The most important theoretical result of the work is the higher rate of growth of the linear approximation of the dependence of the block-average values of the frequency indexes of singular vectors on its number, which was established for the original content, as compared to the non-original content. The significance of the obtained results is that the developed expert method, being a block one, has insignificant computational complexity – O (n2) operations for an nx n image matrix, which makes it promising for working with digital content, in particular with video, in real time. At the same time, the algorithmic implementation of the method made it possible to increase the efficiency of detecting violations of image integrity by reducing type I errors by 2% compared to the best analogue; type II errors remained at the same level. The versatility of the method is confirmed by its high efficiency regardless of the specifics of the perturbation effect, including in conditions of minor disturbances.