Analysis of Image Steganography using Wavelet and Cosine Transforms

Abstract

In this cyber era, protecting data has been a must, one of which is performed by implementing data hiding methods or steganography. It is a technique of hiding data or messages into a media cover, which can be an audio, image, or video file. This paper examines two transformed-based steganography, implementing DCT (Discrete Cosine Transform) and DWT (Discrete Wavelet Transform) to two types of cover images, general and medical. The purpose of this research is to analyze the performance of the embedding process using the low, middle, and low-middle frequency components of the DC coefficient and embed the payload bits into Low-Low (LL), Low-High (LH), High-Low (HL), High-High (HH) of the DWT bands. The experimental results show that, in general, the DCT method has better PSNR (Peak Signal to Noise Ratio) and payload capacity than DWT. Meanwhile, the combined low and middle frequency in the DCT method has a greater PSNR than other components in this transformation, which is 51.6751. Whereas in the DWT technique, LH and HL frequency components show better PSNR values than other DWT band components, which are 43.6939 and 43.5576, respectively.