The paper [2], discusses about authentication of voters in a Remote Voting System (RVS), a new VC scheme is suggested. The current technique, unlike classical VC, is based on the creation of a new matrix utilizing the bitwise XOR operation. Because it uses structural similarity measures like Peak Signal to Noise Ratio (PSNR) and Mean Square Error (MSE) for verification, this authentication system is ideal for RVS. The experimental results show that the suggested approach is computationally efficient and achieves a decent balance of security, storage, and performance.

Proposed Steps:

1) Matrices Creation: The suggested scheme is not intended to be expandable. As a result, one row of the C0 is randomly chosen to encode a white pixel, and one row of the C1 is randomly chosen to encode a black pixel. Then, for each row, one element will is assigned to one share and the other to another. Furthermore, the Hamming weight H = two and the pixel expansion m equals one.

2) Shares Generation Phase: If the pixel color is white, choose one block at random from the two blocks in the codebook that are peer to the white pixel shown in Figure 2, assign one element to share 1 and the other element to share 2. If the pixel is black, choose one block at random, repeat the process that peer to the black pixel shown in Figure 2. When every pixel of the secret image is scanned, the two shared images will be created.

3) Recovery Phase: The original secret image is recovered by stacking the shared images using the bitwise XOR method, as shown in Figure 3.

Advantages and Results Analysis:

- Because the reconstructed images are the same size as the original secret image, the pixel expansion equals one.

- As a result, it saves voters’ and servers’ storage space and allows for quick transmission through public networks, and other communication channels.

- The image created by stacking the shared photographs is accurate and has no information loss. The PSNR and MSE are mathematically expressed in Figure 4.

- Lower variance between the secret image and the reconstructed image with good visual quality is indicated by a greater PSNR and lower MSE. When the PSNR value is equal to and the MSE is zero, the technique gives the best visual quality possible, with no discernible difference between the recovered image and the original secret image.

- Figure 5 demonstrates the MSE and PSNR values obtained between the original and recovered images for two voters: voter 1 provided the original share, whereas voter 2 submitted a suspected share. These findings show the present scheme’s efficiency and effectiveness by approaching typical MSE and PSNR values. As a result, it is an appropriate strategy for RVS authentication.

As a result, the proposed method could cut operational costs and time while still fulfilling the security requirements for high-performance RVSs.

Paper [3] discussed VC techniques that are used for privacy protection. By implementing a cryptographic encryption methodology utilizing pixel shuffling with interchanging their location to generate the ciphered image, it would make it difficult to decipher the image without prior knowledge of the algorithm and the secret key used.

The authors of paper [4] developed an online voting system for an Indian association named “Maharashtra Carrom”. The system incorporates both a CAPTCHA

code and split-image in-share technology, as well as anonymous voting to guarantee private and secure voting process. VC is a safe method that divides a secret image into small fragments. Using two-out-of-two VC, authors propose an Internet voting system, in which users provide a secret image as a password, which is then split into two shares. One share is sent to each user’s email address as a password via VC, while the other is saved on the server. As illustrated in Figure 9, when users log in, they are asked to enter the share password they received through email therefore the system will add share 2, and if the password is correct, a CAPTCHA code will be generated and displayed.

The CAPTCHA code is the image converted into black and white then it will be split into shares. Each share is composed of equal numbers of black and white

pixels. The shares will be merged at the time of voting, and the CAPTCHA will appear if the user is legitimate. Voters determined to be malicious will not be allowed to cast their votes, and the system will immediately log them out. Although the authors claim that using their proposed system will result in making people’s life easier, the system not yet been implemented.

This paper [5] addresses that the voters can vote only if they login to the system with the correct credentials. Share 1 is sent to the voter’s e-mail address before the election and share 2 is provided in the system for the login process throughout the election. Voters will obtain the password to submit their vote by joining share 1 and share 2 using VC. To perform the required services as proposed in the system design stages by [5], the browser must communicate with the HTTP server. Then for users to submit their votes, an e-ballot (gadget used to cast votes in an election) is generated through steganography. The vote is casted using a database server and then encrypted via VC at the application server.

In the proposed system of [6] a connection between the client and server must be established using an algorithm such as in Figure 10.

Afterwards the needed information, such as email address is obtained from the user on the registration page. Then an image is sent to the user to start the subsequent level of the process shown in Figure 11.

It highlights the utilization of VC algorithm to verify, authorize and authenticate user credentials to allow the user to cast their vote. The authors [6] suggested five working modules: initializing the server, sharing password, authenticating, casting the vote, and analyzing the result. The responsible party of the elections must submit the name of the nominees to the host, then the server system will initiate and receives votes. An email attached with a visual key share is sent to the user, while the other share is saved in the database. Authenticating

the sent visual key share with the one stored in the database is done by cross referencing via the server, which produces an image that must be inserted in the prompted page in Figure 12.

If the user is authenticated, they can progress to casting the vote without a chance of forgery. Also, analysis of election results is possible through the server end. This proposed system implements a combination of client, server, database, and cryptography method to ensure security and integrity as illustrated in Figure 13. The aspect of receiving the needed share through email is what makes this method suitable for preventing phishing attacks.

The proposed system aims to guarantee legitimacy, security, and confidentiality of votes casted by users. It is possible by combining biometric information in a multi-party computation, physical characteristics used to assure the identity of a person through electronic devices, with VC [7]. The system takes advantage of multi-party computation features to tally votes. There are four modules in the proposed system that consists of: voter registration, authentication, vote cast and record, and vote count and announcement of the result. In the registration module the algorithm in Figure 14 is used for the authentication and enrollment of vote, which each voter must scan their fingerprint before casting a vote. One

of the shares that contains the biometric minutia is saved in the Voter Identification Card (VIC), while the other is stored in the database. The VIC contains private information belonging to the voter to be used for voting authentication, it serves as a sufficient method for both security and convenience.

The two shares are reassembled to produce the original fingerprint image, which is used to compare with a new fingerprint image obtained from a scanner. If there is a match the voter is legitimate, therefore allowed to vote as in Figure 15.

There will be a voting machine with buttons to record the votes for each nominee, each press by the user generated a signal in binary. However, storing the vote directly in this form is insecure and a new system is proposed by the authors [8] to be more secure. Where “D” is a number greater than sum of number of voters, and “n” is the number of candidates. Let (Rl*d), (l + R2*d), (R3*d),···, and (Rn*D). The database would look like that in Figure 16, and each column symbolizes a vote for each nominee.

Phishing is a prevalent attack targeted on voters using e-voting systems, the severity of damages done by it calls for the use of secure schemes. Due to its importance, several proposed systems were published in the past decade. Most of the methodologies adopted VC as an integral part of its system in addition to some improvements, such as in paper [2] [3] [5] [6] [8]. In paper [2] and [8] a variation of an XOR algorithm was used for verification, which can benefit e-voting systems. However, it can be easily bypassed with proper knowledge. Whereas papers [5] and [6] adopted basic VC schemes, along with email authentication. Implementing a simple combination of authentication methods could pose to be a tremendous security disadvantage. In case the database that stores the image shares becomes compromised, a malicious entity could deceive the authentication. The vantage point in paper [8] pairing fingerprint scanners in the process of registration and authentication for e-voting. This method ensures a lower chance of false authentication, which is an improved security mechanism. As for paper [3] it implemented three VC schemes with two of them discussed anti-phishing in particular expansion less share, and image captcha base authentication.