HIGH CAPACITY IMAGE STEGANOGRAPHY TOLERATING IMAGE COMPRESSION

Abstract

Nowadays, data privacy becomes more important. Most of the existing data protection schemes based on cryptography cannot hide the existence of secret data. Thus, a steganography approach plays an important role to handle this problem. Image steganography can be categorized into two methods which are baesd on spatial domain and transform domain. Methods using spatial domain have high embedding capacity but they are not robust to image compression. On the other hand, methods using transform domain usually work with image compression but have low embedding capacity. In recent image steganography research, many transform domain methods are based on JPEG format which is quite old and its compression efficiency may not be suitable in cloud era. In 2010, Google introduced WebP, a new image format that has better lossy compression than JPEG. In this study, the new image steganography based on WebP is presented. Re-calculation and LSB on middle-frequency method are used. The proposed scheme provides better image quality and smaller file size than the existing JPEG based method at the same capacity. Moreover, not only applying to a grayscale image but also a color image and an alpha channel color image is introduced in this research.