Double Visual Cryptography Using Generalized Tent Map, Rotation, and Image Filtering

This paper introduces a Multi-Visual Cryptography (MVC) system for sharing two color images, where the secrets can be revealed with low computation power using all the shares. The system uses the generalized Tent map as a source of randomness to generate any number of random shares. More specifically,
$(n-1)$
random shares are generated, and then, the
$n^{th}$
share is calculated from the random shares and the secrets using rotations of the shares. In recovery, rotation of the last share recovers the two images based on the angle of rotation. Half the number of pixels is recovered for each secret image, whereas a modified weighted average filter is used to improve the quality of the recovered images significantly. The system does not use halftone images and produces shares of the same size as the secrets without pixel expansion or auxiliary data. The proposed system is efficient, passed several security tests, and is compared to recent works.