Visual image encryption scheme based on vector quantization and content transform

The security issues of outsourced storage make it challenging to protect the privacy of image contents. The noise-like encrypted image generated by traditional image encryption schemes are easy to attract the interest of attackers, courting malicious attacks and analysis on it. In this paper, a novel visual image encryption scheme is proposed to encrypt original image into a visually meaningful camouflaged image, thereby not only protects original images in an encryption way, but also provides an additional visual protection to avoid the risk of being attacked. The original image is first subjected to vector quantization (VQ) encoding, and then visually encrypted by content transform based on discrete wavelet transform (DWTCT) to generate the camouflaged image. The experimental results show that the size of camouflaged image is the same as original image, thereby maintaining the storage efficiency in the outsourced cloud environment. Compared with existing visual image encryption schemes, the proposed scheme shows outstanding performance on both storage and camouflage characteristics of the encrypted images.