On Empirical Evaluation of Partial Encryption for Image Sets

Many professional image databases contain a large number of similar images, which may be stored as a representative image and the differences between the originals and the average. To reduce the heavy computation load of encryption, there are two choices of partial/selective encryption. One is to encrypt a small percentage of every image. The other is to encrypt the average image partially and a tiny portion of each difference image. This paper reports an empirical study on the performance of partial encryption of image sets. In our experiment, the portion to be encrypted is simply hidden from the reconstruction algorithm. No actual encryption algorithm is used. The performance of encrypting different portions of the image set is evaluated by visual inspection and the PSNRs. Hiding the average and portions of the differences provides a better encryption performance with a small fraction of the