QUANTITATIVE ASSESSMENT OF THE QUALITY AND ROBUSTNESS OF SELECTED STEGANOGRAPHIC ALGORITHMS

Abstract

This paper presents the results of a comparative analysis of three widely used steganographic approaches—LSB (Least Significant Bit), DCT (Discrete Cosine Transform), and PVD (Pixel Value Differencing)—applied to embedding textual messages into digital images with varying sizes and resolutions. The main objective of the study is to identify the strengths and weaknesses of each method through the use of quantitative metrics that provide an objective evaluation of the quality and robustness of steganographic algorithms.

Four metrics were selected for the analysis: Mean Squared Error (MSE), Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), and Normalized Cross-Correlation (NCC). These indicators make it possible to quantitatively assess the degree of image distortion as well as the preservation of structural properties after the information-hiding process.

Experimental results demonstrate that the LSB algorithm achieves the highest image quality with virtually imperceptible visual distortions. In contrast, the DCT method exhibits more noticeable degradation, although it may be suitable in scenarios where a reduction in image quality is acceptable. The PVD algorithm showed the shortest execution time and provided a balanced combination of quality, robustness, and performance, indicating its potential suitability for systems focused on high-speed data processing.

The conducted study confirms the effectiveness of using a set of quantitative metrics for comprehensive comparison of steganographic methods. The proposed approach enables a deeper analysis of algorithm behavior under various conditions and forms a foundation for the development of adaptive steganographic solutions aimed at enhancing information security.