Generating random grid-based visual secret sharing with multi-level encoding

Most visual secret sharing (VSS) schemes are employed for the secret sharing of binary secret images. In these schemes, the share appears as a noise-like image, such that no information about the secret image can be obtained from an individual share. The recovered secret message can only be decrypted by the human eye following rendering using two different light transmissions. However, these schemes do not apply well to gray-level secret images because the halftone region of gray-level secret images cannot be adequately recovered. In this paper, we propose a VSS scheme with multi-level encoding based on a generalized random grid, which can retrieve the secret in a multi-level secret image with a high visual quality using the XOR operant during decryption. The proposed method produces a share of the same size as the original secret image without pixel expansion and does not require a codebook in the encoding process.