Media Steganography with CNN & Blockchain for Secure Transfer

Secure data transfer through internet has become highly important because of the loopholes and vulnerabilities of the existing data transfer systems. Steganography is an effective technique to achieve this goal, where data is hidden within multimedia files to prevent detection. In our implementation, we have incorporated a technique, CNN-based image and video steganography, which leverage Blockchain technology for efficient storage and transfer of hidden data. Our technique involves compressing and dispersing the original secret image pixels across all of the available bits of the container image, in contrast to traditional steganography methods which encrypt the secret message by embedding it within the least significant bits (LSBs) of the container image. CNN (Convolutional Neural Network) can help achieve two goals in steganography: to increase security and increase hiding capacity. CNNs can be used to identify the best areas for hiding data, which increases hiding capacity and it becomes complex for Steganalysis programs to detect the hidden data. The proposed methodology also includes a decentralized cloud system which uses Blockchain technology to store and transmit private media files providing decentralization, transparency and tamper-evidence.