An adaptive algorithm for embedding information into compressed JPEG images using the QIM method (2012.08742v1)

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.