Description

Steganography is the art and the science of hiding information into covert channels, so as to conceal the information & prevent the detection of the hidden message. To improve the security performance of JPEG Steganography, there are two approaches are used. In Matrix Embedding, less modification needs to the cover image for embedding the secret bits. Another one is channel selection; it is used to find the dct coefficients for embedding the secret bits. In this paper, the channel selection rule (to find the coefficients, which introduce minimal detectable distortion for data hiding) is applied, three factors are considered for the channel selection rule. i.e.) Perturbation Error (PE), the Quantization step (QS) & the magnitude of quantized DCT coefficients to be modified (MQ). To find the coefficients, the image are splited into equal sizeable blocks & apply the two dimensional DCT to the image by using these factors. The magnitude of the DCT coefficients with relatively large magnitude, & low frequency. After finding the dct coefficients, the channel selection rule is based on the modified matrix encoding algorithm; it allows more than one embedding into each coefficient blocks. After, that embedding process, there is no difference between the cover image & the stego image (the image which contains more than one embedding changes into each coefficient blocks). Finally, it provides the higher security performance for JPEG Steganography for hiding the data.