Abstract
Steganography is a method of data hiding. It is more and more important while the Internet communication grows up. Now, many steganographic algorithms or software have already been developed including image hiding techniques for JPEG images. JPEG is the most common image format used by digital cameras and transmitted in the Internet. This thesis proposes two methods to increase the capacity of F5 embedding algorithm. The first method is a pre-processing method, we randomly modifies the qDCT coefficients before the F5 embedding algorithm to get more nonzero qDCT coefficients to embed message. In the second method, we slightly modify the end of block (EOB) information in a JPEG image to increase the capacity of embedding after the F5 embedding algorithm. We change a qDCT coefficient on each qDCT 8×8 block to embed 3 more bits of message. Although the PSNR value is slightly decreased, our proposed algorithm significantly increases the capacity of embedding compared with many classical steganographic algorithms.