An efficient visually meaningful image encryption using Arnold transform

Image encryption is a quite well-developed area of multimedia security which helps to protect the confidentiality of authorized users' image content by converting it into an incomprehensible form. Most of the image encryption algorithms will generate encrypted images with noise-like or texture-like contents. Hence it's easy to distinguish natural images from encrypted images. The chances of attacks and image analysis by malicious people are more on images with noise-like or texture-like contents. This paper presents an efficient visually meaningful image encryption algorithm that uses integer discrete wavelet transform(DWT) and Arnold transform. The central idea of proposed scheme is to use a reference image to hide the original image in its encrypted form. Arnold transform has been used to transform the encrypted original image content before inserting it into the wavelet sub-bands of the reference image. The final encrypted image will look similar to the reference image, and the original image can be extracted and decrypted by an authorized person who has the valid decryption and data hiding key.