Performance Analysis of Image Steganography Based on Least Significant Bit and Pseudorandom Encoding Techniques

Abstract

The increasing demand for secure communication over public networks has intensified research on information hiding techniques capable of protecting sensitive information from unauthorized access. Image steganography achieves this objective by concealing secret information within digital images while preserving the visual appearance of the cover image. Among the numerous image steganographic techniques, the Least Significant Bit (LSB) method is widely adopted because of its simplicity, high embedding capacity, and low computational complexity. However, its sequential embedding mechanism makes it susceptible to steganalysis and statistical attacks. To address this limitation, pseudorandom embedding techniques distribute secret information randomly throughout the cover image using a shared secret key, thereby improving security and imperceptibility. This paper presents a comparative performance analysis of the LSB and Pseudorandom Encoding techniques for secure image steganography. Both techniques were implemented in MATLAB using identical cover images and secret messages to ensure an unbiased evaluation. Performance was assessed using Peak Signal-to-Noise Ratio (PSNR), Mean Square Error (MSE), Signal-to-Noise Ratio (SNR), embedding capacity, and visual image quality. Experimental results demonstrate that although both techniques effectively conceal secret information with negligible perceptual degradation, the Pseudorandom Encoding technique consistently produces higher PSNR values and lower distortion than the conventional LSB method. Furthermore, the random distribution of embedded data significantly improves resistance to unauthorized detection while maintaining comparable embedding capacity. The findings confirm that pseudorandom embedding provides a more secure and robust alternative to sequential LSB steganography for digital image information hiding.