An Intelligent and Hybrid Method of Combining Spatial and Frequency Representation for Digital Image Steganography

Data security becomes important concern in this internet world. Lots of internet users are losing their conceal data when they are communicating it through the internet. So, there is a demand to adopt secured communication technique to ensure the safety of users' data. One such technique is steganography. Steganography is the practice of embedding conceal information into another skin data which is not easy to recognize the conceal information by unintended person. The conceal and skin data may be of any format. In this proposed method, conceal image is hided into a skin image by using hybrid method where steganography is performed by using both spatial as well as frequency representation as a hybrid method. By using the proposed method, the secureness of the conceal image is assured by three processes whereas in conventional method, only one process is used. Initially, in spatial representation, binary values of conceal image pixels are hidden by mapping suitable locations in the skin image pixels. Then, the identified suitable locations are extracted as coefficients which converted to gray codes. Finally, the gray codes are embedded into the skin image in its frequency representation which is transformed using Haar DWT. The resultant stego image is communicated to the recipient side where the conceal image is recovered back from stego image by inverse DWT and mapping coefficients. The experimental results show better performance parameters like entropy, cross correlation, MSE, PSNR and SSIM values compared to that of conventional LSB method.