An adaptive algorithm for embedding information into compressed JPEG images using the QIM method

TL;DR

This paper introduces an adaptive steganographic algorithm for embedding information into JPEG images using variable quantization steps, enhancing security against statistical steganalysis.

Contribution

It presents a novel adaptive QIM-based embedding method that adjusts quantization per image block to improve steganographic security.

Findings

Increased resistance to statistical steganalysis.

Effective embedding with minimal image distortion.

Demonstrated improved security in experimental tests.

Abstract

The widespread use of JPEG images makes them good covers for secret messages storing and transmitting. This paper proposes a new algorithm for embedding information in JPEG images based on the steganographic QIM method. The main problem of such embedding is the vulnerability to statistical steganalysis. To solve this problem, it is proposed to use a variable quantization step, which is adaptively selected for each block of the JPEG cover image. Experimental results show that the proposed approach successfully increases the security of embedding.