Visual cryptography scheme for secret color images with color QR codes

https://doi.org/10.1016/j.jvcir.2021.103405Get rights and content

Highlights

  • This paper proposes two schemes that can completely restore the secret image.

  • The QR code is used as the share in the proposed scheme.

  • This paper proposes two pixel-non-expansible schemes.

Abstract

Visual cryptography scheme can divide the secret image into several shares. It can be used to enhance the secure transmission of the secret image on the Internet. Most schemes cannot fully restore the secret color image and generate meaningful shares. This paper proposes two new schemes by using color XOR to solve these problems. The first proposed scheme generates meaningless shares. The second proposed scheme can generate n1 meaningful shares and a meaningless share. Based on the first proposed scheme, n1 shares are modified to color QR codes in the second proposed scheme. These color QR codes can be decoded by the general decoder instead of the standard decoder. All shares are performed the color XOR to restore the secret color image completely. This paper uses some experiments to test two proposed schemes. Experimental results show that the two proposed schemes are feasible.

Introduction

Visual cryptography scheme (VCS) is a technology that can ensure the secure transmission of the secret image on the Internet. It was proposed by Naor and Shamir in 1994 [1]. The secret image is divided into several shares by VCS. Specific number shares can be performed in the decoding operation to restore the secret image. When a small number of shares (the number is less than the specific number) are obtained by the receiver, they cannot restore the secret image. The secret information is not obtained. Every share is a key that can enhance the secure transmission of the secret image on the Internet. If the number of obtained shares is less than the given threshold by the attacker, they cannot also obtain the secret image.

The VCS can enhance the secure transmission of the information on the Internet [2], [3], [4], [5], [6], [7]. Besides, some other methods can also be done, such as watermarking [8], [9], [10], [11], [12], information hiding [13], [14], [15], [16], [17], [18], the public key [19], [20], [21], the private key [22], [23]. Watermarking, information hiding, the public key and the private key all have the same disadvantage. When the decoding method is known by the attacker, they can easily decode the secret information. The VCS can generate several shares. Every share is a key. If the attacker knows the decoding method and obtains insufficient quantity shares, they still cannot decode the secret information. The VCS mainly is used to ensure the secure transmission of the secret image in the network.

The VCS can share binary images, grayscale images and color images. It is mainly divided into two types. One type is the pixel-expansible scheme and another type is the pixel-non-expansible scheme. The pixel-expansible scheme uses the basic matrices to achieve the VCS. Naor and Shamir proposed VCS by using two basic matrices [1]. Every share is a meaningless binary image, and is a key. The attacker cannot reveal the secret image from any share. All shares are printed onto transparent films and overlaid to restore the secret image. This decoding method is also called OR. When the number of shares is less than the threshold, the secret image cannot be restored. Hou proposed a new VCS for color images by using the subtractive model [24]. Luo et al. used halftone technology to achieve VCS for color images [25]. Grayscale images and color images own continuous tone. They are hard to be shared by VCS directly. These continuous tone images need to be converted to halftone images by using halftone technology.

The size of every share is larger than the original secret image in the pixel-expansible scheme. Someone proposed pixel-non-expansible VCS to solve this problem. Shyu proposed a new VCS to generate pixel-non-expansible shares by using random grids [26]. Lin et al. proposed a new VCS which can generate pixel-non-expansible shares by Hilbert-curve and two queues [27]. Fang proposed a reversible VCS by using random grids and halftone technology [28]. Bakshi et al. achieved the VCS without pixel expansion by inserting some information in the region of non-interest [29]. These schemes belong to the pixel-non-expansible schemes. The size of shares and the restored image is the same as the secret image. Generated shares are meaningless by most schemes. A new scheme that can generate meaningful shares needs to be proposed.

Lou et al. proposed a new VCS that can generate meaningful shares by using halftone technology [30]. Abd El-Latif et al. proposed a new VCS for the binary image by using random grids, error diffusion (ED) and chaotic permutation [31]. Thomas et al. achieved VCS to generate meaningful shares by halftone technology [32]. These schemes cannot fully restore the secret image. This paper proposes two new schemes to share the secret image by using color XOR (CXOR) to solve this problem. Two proposed schemes can fully restore the secret image. The first proposed scheme can only generate meaningless shares. The QR code can also be used as the share to share the secret image [33], [34]. The second proposed scheme can generate n1 meaningful shares and a meaningless share. These n1 shares are all color QR codes. All color QR codes can be decoded by the general decoder instead of the standard decoder.

The main contributions of this paper are summarized as follows:

  • This paper proposes two new visual cryptography schemes for the secret color image. Two proposed schemes can fully restore the secret image. The quality of the restored image is enhanced in two proposed schemes.

  • These n1 shares are meaningful images (color QR codes) and a share is a meaningless image in the second proposed scheme. These color QR codes can be decoded by the general decoder instead of the standard decoder.

  • Two proposed schemes can generate pixel-non-expansible shares. The size of every share is the same as the original secret image. Two proposed schemes belong to pixel-non-expansible schemes in this paper.

The remaining structure of this paper is: Section 2 introduces the preliminaries, including VCS, QR code, ED, the brightness of the color and CXOR. Two proposed schemes are described in Section 3, including the encoding process and decoding process. Section 4 exhibits all experimental results. The conclusion is given in Section 5.

Section snippets

Preliminaries

This section mainly introduces preliminaries. Section 2.1 describes the concept of the VCS and two conditions of the VCS. The structure and basic knowledge of the QR code are introduced in Section 2.2. Section 2.3 shows a halftone technology ED. The brightness of the color is described in Section 2.4. Section 2.5 introduces the operation of CXOR.

The proposed scheme

The process of the proposed scheme is shown in Fig. 5. The color image C is not encrypted by the proposed scheme directly. It needs to be converted to a halftone image H which is the secret image. H can be divided into n shares by the proposed scheme. All shares are performed CXOR to restore the secret image. The restored image H is the same as H. The proposed scheme is the (n,n)-threshold scheme. The size of shares is the same as the secret image. The proposed scheme is pixel-non-expansible.

Experiments

This section mainly introduces all experimental results. The generated QR code follows the Zxing library [43]. The designing of the color QR code follows ISO [44]. The software is Matlab2016b.

Conclusions

This paper proposes two visual cryptography schemes that can fully restore the secret color image. Two proposed schemes belong to (n,n)-threshold schemes. The first proposed scheme generates meaningless shares. The second proposed scheme can generate n1 meaningful shares and a meaningless share. These n1 meaningful shares are color QR codes that can be decoded by the general decoder instead of the standard decoder. Two proposed schemes are pixel-non-expansible. Although shares are color QR

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (44)

  • HouY.-C.

    Visual cryptography for color images

    Pattern Recognit.

    (2003)
  • LuoH. et al.

    Color transfer in visual cryptography

    Measurement

    (2014)
  • ShyuS.J.

    Image encryption by random grids

    Pattern Recognit.

    (2007)
  • BakshiA. et al.

    Secure telemedicine using RONI halftoned visual cryptography without pixel expansion

    J. Inf. Secur. Appl.

    (2019)
  • LouD.-C. et al.

    A novel authenticatable color visual secret sharing scheme using non-expanded meaningful shares

    Displays

    (2011)
  • Abd El-LatifA.A. et al.

    A new meaningful secret sharing scheme based on random grids, error diffusion and chaotic encryption

    Opt. Laser Technol.

    (2013)
  • NaorM. et al.

    Visual cryptography

  • YanX. et al.

    Reversible image secret sharing

    IEEE Trans. Inf. Forensics Secur.

    (2020)
  • LiuT. et al.

    Color visual secret sharing for QR code with perfect module reconstruction

    Appl. Sci.

    (2019)
  • YanB. et al.

    Improving the visual quality of size-invariant visual cryptography for grayscale images: An analysis-by-synthesis (AbS) approach

    IEEE Trans. Image Process.

    (2018)
  • WangL. et al.

    Flip extended visual cryptography for gray-scale and color cover images

    Symmetry

    (2021)
  • HuangP.-C. et al.

    Efficient QR code authentication mechanism based on Sudoku

    Multimedia Tools Appl.

    (2019)
  • Cited by (0)

    This paper has been recommended for acceptance by Radeva Petia.

    View full text