An Overview of Image Steganography Techniques: Historical Development, Methodologies, and Evaluation Criteria

Abstract

This paper provides a detailed overview of image steganography, tracing its historical roots and examining the development of modern techniques. Steganography, derived from the Greek words "stegos" (cover) and "graphia" (writing), refers to the practice of concealing information within a host medium. Historically, steganographic techniques included tattooing hidden messages and using invisible ink, but the digital age has introduced new possibilities for secure data embedding, particularly within images. With the advent of the internet, the need for robust steganographic techniques has increased, leading to advancements in both spatial and frequency domain methods, including Least Significant Bit (LSB) replacement, Discrete Cosine Transform (DCT), and wavelet-based techniques. This paper also reviews evaluation criteria such as Peak Signal-to-Noise Ratio (PSNR), Mean Squared Error (MSE), and Structural Similarity Index Measure (SSIM) to assess the effectiveness of various steganographic methods in achieving high security, capacity, and imperceptibility. Challenges such as balancing these criteria, developing multi-criteria evaluation tools, and optimizing embedding processes are highlighted. This review aims to provide insights into current advancements, limitations, and future directions for research in image steganography.