Steganography is the art of secret writing, where one message, file or image is hidden inside another to hide its existence. There are many different algorithms that use different techniques and implementations to hide data in various forms of digital media. This article will focus on the comparison of two image steganography algorithms, namely F5 algorithm and YASS (Yet Another Steganographic Scheme). The article will first provide a brief overview of each algorithm and their implementations with their advantages and disadvantages. We will then compare algorithms based on type, overlay image formats, embedding channels, use of stego keys, robustness, and security. Both the F5 and YASS algorithms were developed to embed secret messages within the JPEG image format. This image format is an example of lossy compression that makes use of the discrete cosine transform (DCT) (Cabeen & Gent, 1998: 1). Since both the F5 and YASS algorithms embed messages in JPEG images, we first need to provide an overview of what happens during JPEG compression. When a JPEG image is compressed, the color model is first transformed from the RGB model to the YCBCR model. This transformation is not always necessary but allows for higher compression rates for the same level of reliability (Fridrich, 2010: 22). Each image channel is then divided into 8 x 8 pixel blocks as shown in Figure 1 below. Chroma signals (CBCR components) could be subsampled prior to this chunking as the human eye is less sensitive to changes in chrominance than luminance (Fridrich, 2010: 23). Next, the pixel values of each block are moved from the range [0.255] to a new range [-128, 127] that center... middle of the sheet... concerns the security performance of an algorithm. They compared the highest detection accuracy rates achieved by the 4 detection algorithms for F5 and YASS and showed the accuracy rates along with the bits per non-zero DCT coefficient for each algorithm. The diagram can be seen below in Figure 6. Conclusion Both F5 and YASS algorithms are examples of JPEG steganography and each has its own strengths and weaknesses. For both algorithms, the researchers proposed different ways in which embedded data could be detected using the algorithms. Since their appearance, many variations of the two algorithms have been created that aim to improve aspects such as efficiency, robustness and embedding security. Unfortunately, there will always be a trade-off between embedding capabilities, robustness, and security/undetectability of steganographic algorithms.
tags