International Scientific Journals
of Scientific Technical Union of Mechanical Engineering "Industry 4.0"

Steganography aims to conceal confidential information within digital media while minimizing perceptual and statistical distortions. Among image steganographic techniques, adaptive embedding methods have demonstrated superior resistance to steganalysis by exploiting image regions with high local complexity. One effective approach utilizes the Canny edge detection algorithm to identify suitable embedding locations. However, the effectiveness of such methods depends not only on embedding capacity but also on comprehensive security evaluation. This paper presents a systematic framework for assessing the security of an adaptive Canny-based steganographic method. Multiple objective metrics, including Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), Mean Squared Error (MSE), histogram similarity, entropy variation, correlation coefficients, and resistance to statistical steganalysis are analyzed. The proposed evaluation framework demonstrates that edge-based adaptive embedding maintains high visual fidelity while significantly reducing detectability compared with conventional least significant bit (LSB) techniques.

Once the steganography was invented, it still remains one of the most effective ways to hide information in publicly accessed data sources. Many applications of different steganography techniques were known since WWII, but there are no yet any good approach to detect whether there is a hidden message in some data source (image etc.) while it is not expected to be found. This may also cause a fundamental security breach in corporate security systems. In the era of Facebook, Instagram and many other social networks with a huge amount of images and photos it seems obvious to use steganography applications for the purpose of security breach. A saboteur or a spy can use any image from social networks to disclose corporate sensitive data or to exchange hidden messages. That is why steganography detection should be considered as a necessary tool to prevent security breaches and sensitive data leaks. This research is focused on analysis of current approaches in problem of steganography detection. Mainly the approaches to detect steganography in images were discussed.