Comparative Analysis of Visual Cryptography Using Codebook and Random Grid Visual Cryptography

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 12 Issue: 06 | Jun 2025 www.irjet.net p-ISSN: 2395-0072

Comparative Analysis of Visual Cryptography Using Codebook and Random Grid Visual Cryptography

Anil B Alde1 , Vikas T. Humbe2

1School of Technology, S.R.T.M.U.N, Sub-Campus, Peth, Latur, Maharashtra, India

2School of Technology, S.R.T.M.U.N, Sub-Campus, Peth, Latur, Maharashtra, India

Abstract - Visual Cryptography (VC) is a cryptographic technique that splits a secret image into shares, revealing no information individually but reconstructing the image when stacked together. This paper compares two VC methods: Random Grid Visual Cryptography (RG-VC) and Visual Cryptography using Codebook (VC-CB). We analyze their principles, implementation, security, efficiency, and applications. Our findings show that RG-VC offers simplicity and perfect security, while VC-CB provides scalability and flexibility. These insights help inselectingthe appropriate VC method for various applications and further research work.

Key Words: VisualCryptography(VC),RandomGrid Visual Cryptography (RG-VC), Visual Cryptography usingCodebook(VC-CB),Security

1. INTRODUCTION

Visual Cryptography (VC) is a powerful cryptographic technique for secure image sharing, with applications in securecommunications,authentication,andwatermarking. VC divides a secret image into multiple shares that individually reveal no information [1]. However, when shares are superimposed one above the other, the secret imageisrevealed[2]asfiguredoutinFig.1.

This paper compares two VC methods: Random Grid VisualCryptography(RG-VC)andVisualCryptographyusing Codebook (VC-CB). We analyze their principles, implementation,security, efficiency, andapplications. The paper is structured as follows: Section 2 discusses the principles of VC-CB and RG-VC, Section 3 compares their features, Section 4 details their algorithms, Section 5

presentsexperimentalresults,andSection6concludeswith keyinsights.

1.1 VC-CB



Naor and Shamir [3] presented a VC-CB in 1994, in whichacodebook,apredefinedsetofpatternsorcodes isusedtoencodethesecretimageintoshares.

 Eachsharecontainsaportionofthesecretimage,and whencombinedaccordingtothecodebook,theoriginal imagecanbereconstructed.

 UnlikeRandomGridVisualCryptography,whichrelies on random patterns, this approach uses a predeterminedsetof codes toencodeanddecodethe secretimage.

 Thesecurityofthismethoddependsonthesecrecyof the codebook. If an adversary gains access to the codebook,theycaneasilyreconstructtheoriginalimage fromtheshares.

1.2 RG-VC

 The encryption of pictures by two random grids was firstintroducedbyKafriandKerenin1987[4]andwas neglectedfortwodecadesuntil 2007[5]inwhichthe secret image is divided into multiple shares, each of whichcontainsrandomgridsofblackandwhitepixels. Thesesharesaredistributedamongparticipants.

 Torevealthesecret,theparticipantssuperimposetheir sharesontopofeachother.Onlywhenthesharesare overlaidinthecorrectalignment,doesthesecretimage becomevisible.Individually,thesharesdonotprovide anyinformationabouttheoriginalimage.

 This technique relies on the randomness of the grid patterns to achieve secrecy. The random grids are designedinsuchawaythattheyrevealthesecretwhen overlaidcorrectlybutremainmeaninglessotherwise.

2. PRINCIPLES OF VC-CB AND RG-VC

2.1 VC-CB

VisualCryptographyusingCodebook(VC-CB)relieson predefined codebooks to encode and decode shares. The secretimageisdividedintoblocks,andeachblockismapped toacodeinthecodebook.Sharesaregeneratedbasedonthe codesinthecodebook[6].VC-CBinwhich2-out-of-2abasic

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 12 Issue: 06 | Jun 2025 www.irjet.net p-ISSN: 2395-0072

matrix,calledacodebookwithallofthecodewordbeingof thesize2X2subpixelsisemployedasdepictedinFig.2.But designingthematrixforencryptionisalsoadifficulttask[7]. Thiscodebookencodesasecretimagetogeneratetwosecret imagescalledshares.

Fig -2: CodebookofVisualCryptography

InFig.2whereVisualCryptographyincludestwogroups,a whitegroupwithsixcasesandablackgroupwithsixcases.If acertainpixelofthesecretimageiswhite,share1andshare 2areassignedcodewordbyrandomlyselectingoneofthesix casesinthesharegroup.Onthecontrary,ifthesecretpixelis black,thecodewordareselectedfromtheblackgroup.The sizeofthesharedimagesShare1andShare2becomelarger thanthatofthesecretimage[8][9].

2.2 RG-VC

Toovercomethedrawbackofthecodebook,researchers conferred on another VSS scheme by the random-grid algorithm(RGVSS),whichwasfirstpresentedbyKafriand Kerenin1987[4],butneglectedfortwodecadesuntil2007 [5].RandomGridVisualCryptography(RG-VC)operatesby dividingthesecretimageintoagridofrandompixels.Each shareconsistsofagridwithrandomblackandwhitepixels. Whenstackedtogether,thesharesvisuallyrevealthesecret image.InRG-VCeachpixelistreatedasagridandthecolour ofthegridisrandomlyassignedasillustratedinFig.3

DuringtheencryptionprocessofRG-VC,theGridR1colour ofthefirstshareisrandomlyassignedandtheGridR2colour ofshare2iseitherthesameorcomplementarydependingon thecolourofthecorrespondingsecretpixel.Shyu[5]extends the kafri’s scheme to propose three algorithms offering differentlighttransmissionswhicharedemonstratedinFig. 3. Ifthe pixel of the secret image is white,thenR1andR2 havethesamecolour.Besidesthat,ifthepixelofthesecret imageisBlack,thenR1andR2arecomplementaryincolour. Thisgenerates100%blackinblackareasand50%blackin white areas when the two shared images are stacked one aboveanother.

InModel2,ifthepixelofthesecretimageisWhite,thenthe coloursofR1andR2arethesamecolouranditgeneratesthe samecolourwhenthesesharesareoverlapped.Ifthepixelof thesecretimageisBlack,thenthecolourofR2israndomly assigned. This generates 75% black in the black areasand 50% black in the white areas when the two shares are overlappedoneabovetheother[3].

InModel3,ifthepixelofthesecretimageiswhite,thenthe colourofR2israndomlyassigned;besidesthat,ifthepixelof asecretimageisblackthenR2iscomplementaryincolourto R1.Thisgenerates100%blackintheblackareasand75% blackinthewhiteareawhensharesarestackedtogether.

Inallthreemodels,nomatterwhatthecolourofthesecret pixelis,theprobabilityofitbeingblackinthesharedimageis 1/2. This ensures the security of the secret image. The contrastsproducedbythethreemodelsare50%,25%,and 25%, which are sufficient for an observer to identify the confidentialinformationinthestackimagewiththenaked eye.

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 12 Issue: 06 | Jun 2025 www.irjet.net p-ISSN: 2395-0072

3. DISTINGUISH FEATURES BETWEEN

VC-CB AND RG-VC

TherearesomanydistinguishablefeaturesbetweenRandom Grid Visual Cryptography and Visual cryptography using Codebook which are helpful for further research work as showninTable1.

Table - 1: FeaturesofVC-CBandRG-VC

Feature VC-CB RG-VC

Nature of Shares Structured patterns fromacodebook

Security Dependsonthesecrecy ofthecodebook

Efficiency Higher computational andstorageoverhead

Applications Multi-party authentication, secure printing

Flexibility / Complexity Depends on codebook design

Random grid Patterns

Perfectsecuritywith randomgrids

Highefficiency

Secure image sharing, authentication

Flexible in grid size andcomplexity

3.1 Nature of shares implementation

VC-CB: Invisualcryptographyusingthecodebooktechnique, shares contain structured patterns corresponding to codewordsfromacodebook.VC-CBimplementationrequires creating and managing codebooks, encoding and decoding blocks, and maintaining codebook security. It may involve complexalgorithmsandadditionalcomputationaloverhead [10]

RG-VC: Inrandomgridvisualcryptography(RG-VC),shares contain random grid patterns resembling noise. ImplementationofRG-VCinvolvesrandomlyassigningblack andwhitepixelstogrids.Itisrelativelystraightforwardand computationallyefficient[4]

3.2 Security Analysis

VC-CB: Itssecuritydependsonthesecrecyofthecodebook.If thecodebookiscompromised,thesecurityoftheschemeis compromised. Careful management of the codebook is essentialtomaintainsecurity[11].

RG-VC: Itoffersinherentsecurityduetotherandomnature of share generation. It provides perfect security, where knowledge of one share reveals no information about the secretimage[1].

3.3 Efficiency

VC-CB: VC-CBmayhavehighercomputationalandstorage overhead due to codebook management and

encoding/decoding processes.It maynot beas efficientas RG-VCincertainscenarios[12].

RG-VC: RG-VC is efficient in terms of computation and storage. It requires minimal processing overhead and producesshareswithlowcomplexity[13]

3.4 Applications

VC-CB: VC-CBissuitableforapplicationswherescalability, flexibility,andmanageabilityareimportant.Itiscommonly used in multi-party authentication, secure printing, and digitalwatermarking[11].

RG-VC: RG-VC is suitable for applications requiring simplicity, efficiency, and perfect security. It finds applicationsinsecureimagesharing,authenticationschemes, andwatermarking[13]

3.5 Complexity

VC-CB: Implementationmaybemorecomplex,asitinvolves creatingand managinga codebook.Generatingsharesand reconstructingtheoriginalimagemayalsorequireadditional computationalsteps[14].

RG-VC: Implementation is relatively straightforward, as it primarily involves generating random grid patterns. No complexencryptionalgorithmsarerequired[15]

3.6Practicality

VC-CB: Shares may be easier to manage and distribute, as they are generated based on a predefined codebook. However,theneedtokeepthecodebooksecureaddsanextra layerofcomplexity[16]

RG-VC: Sharesareeasytodistributeastheyconsistofsimple gridpatterns.However,aligningthesharescorrectlytoreveal the secret imagemay require careful manual alignmentor additionalcomputationaltechniques[17]

3.7Flexibility

VC-CB: Flexibility depends on the design of the codebook. Adding or modifying codes in the codebook may require updatestothesystem[18]

RG-VC: Offersflexibilityintermsofthesizeandcomplexityof the grid patterns. Different sets of random grids can be generatedfordifferentimages[4]

3.8 Comparative Evaluation

VC-CBandRG-VCofferdistinctadvantagesandtrade-offs. VC-CBoffersscalability,flexibility,andmanageability,making itsuitablefordiverseapplicationsOntheotherhand,RG-VC excelsinsimplicity,efficiency,andsecurity.Itissuitablefor scenarioswhereperfectsecurityisparamount[11].

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 12 Issue: 06 | Jun 2025 www.irjet.net p-ISSN: 2395-0072

The choice between VC-CB and RG-VC depends on the specificrequirementsoftheapplication,includingsecurity level,computationalresources,andeaseofimplementation.

4. ALGORITHMIC DIFFERENCE BETWEEN VC-CB AND

RG-VC

4.1 Algorithm of VC-CB

Input: Loadsecretimage

Pre-processing:

 Converttheloadedimageintoabinaryformat(Blackand White)usingathresholdof0.5iftheimageisnotalready Binary

Input shares:

 Getthedimensions(rowsandcolumns)ofthesecretimage. Initializethesharesmatrices(‘share1’,‘share2’)having thesamedimensionsasthesecretbinaryimage.

Codebook Definition:

 Defineacodebookthatspecifiesthepixelpatternsfor whiteandblackpixels.

 Codebookfor2-out-of-2schemes:

o Whitepixel:‘[[1,0],[1,0]]’or‘[[0,1],[0,1]]’

o Blackpixel:‘[[1,0],[0,1]]’or‘[[0,1],[1,0]]’

Generate Shares:

 Foreachpixelinthesecretimage: If thepixeliswhite(pixelvalue=1):

Randomlyselectapatternforawhitepixelfromthe codebook Ifthepixelisblack(pixelvalue=0):

Randomlyselectapatternforablackpixelfromthe codebook.

 Assigntheselectedpatterntothecorrespondingpositions inthesharematrices.

Output:

 DisplayShare1andShare2.

 RevealanddisplaysecretimagesbysuperimposingShare1 andShare2.

4.2 Algorithm of RG-VC

Input: Loadthesecretbinaryimage.

Pre-processing:

 Converttheloadedimageintoabinaryformat(blackand white)usingathresholdof0.5

Generate Shares:

 InitializetwomatricesShare1andShare2ofthesame sizeasthebinaryimage.

For white pixels (pixel value = 1):



Generatetworandomsharesbydefiningsets1a=[0, 1]ands1b=[0,1].

 Assign the first element of each share to the correspondingpixellocationinShare1andShare2.

For black pixels (pixel value = 0):

 Generatetworandomsharesbydefiningsets0a=[1, 0]ands0b=[0,1].

 Assign the first element of each share to the correspondingpixellocationinShare1andShare2.

Output:

 DisplayShare1andShare2images.

 Reveal and display secret images by superimposing Share1andShare2oneaboveanother.

5. EXPERIMENTAL RESULTS

The experimental results of Visual Cryptography using Codebook(VC-CB)inwhichsharesaregeneratedbasedon pre-definedpatternsortemplatesstoredinacodebook.Each templatecorrespondstoaspecificcombinationofblackand white pixels that encode a part of the secret image. To generateshares,thesecretimageisdividedintoblocks,and foreachblock,theappropriatetemplatefromthecodebookis selected and XORed with it to generate a lossless secret image.

Random Grid Visual Cryptography(RG-VC) is another methodology that helps to overcome all the drawbacks of traditional Visual Cryptographysuchas expansionin pixel andlowcontrast[19]andalsogeneratesalosslessresultant secretimageatthereceiversideasthatofVC-CBasshownin Fig.4,Fig.5andFig.6.

Volume: 12 Issue: 06 | Jun 2025 www.irjet.net p-ISSN: 2395-0072

The main difference between RG-VC and VC-CB lies in their complexity, efficiency (generation time), share size, storagerequirementandcomputationalresourceswhichare describedasfollows

5.1 Complexity

RG-VCgeneratessharesusingarandomgridofblackand whitepixels.Thecomplexityprimarilydependsonthesizeof the secret image (dimensions M X N), resulting in a time complexity of O (M X N). This approach involves straightforwardoperationsforsharegeneration,makingit relativelysimple.

VC-CBgeneratessharesbasedonpre-definedpatternsor templatesstoredinacodebook.Thecomplexitydependson thesizeofthecodebook(K)andthenumberofblocksinthe secretimage(B).Therefore,thetimecomplexityisO(KXB). Generation of shares using templates from the codebook involvesadditionallookupoperationsandXORoperations, which may result in a higher computational complexity comparedtoRG-VC.

5.2 EFFICIENCY (GENERATION TIME)

RG-VC typically involves simpler operations for share generation, which may result in faster generation times comparedtoVC-CB,especiallyforsmallerimages.However, forlargerimages,thetimecomplexityofRG-VC(O(M×N)) couldbecomealimitingfactor.

VC-CB:VC-CBcanbemoreefficientintermsofsharesize andstoragerequirementssinceitgeneratessharesbasedon predefined templates. By selecting appropriate templates from the codebook, VC-CB can achieve a more compact representationofsharescomparedtoRG-VC.However,the initial setup involving the creation and storage of the codebookmayrequireadditionalcomputationalresources.

From above Fig. 7, it is clear that RG-VC comparatively require half of the execution time as that of Visual Cryptography using codebook

5.3 Storage requirement

VC-CBcanpotentiallyoffermorecompactrepresentations ofshares,leadingtolowerstoragerequirementscompared withRG-VC,especiallywhendealingwithalargenumberof Shares[20]

5.4 Security and Flexibility

VC-CBmayprovidemoreflexibilityintermsofcontrolling share properties and security levels by designing and managingthecodebookappropriately.

5.5

Computational resources

VC-CBmayrequiremorecomputationalresourcesinitially forthecreationandstorageofthecodebook.However,once the codebook is generated,share generation may be more efficientcomparedtoRG-VC,particularlyforlargerimages.

6. CONCLUSION

This paper provides a comprehensive comparison of two Visual Cryptography techniques: VC-CB and RG-VC. RG-VC offers simplicity, efficiency, and perfect security, makingit suitableforsecureimagesharingandauthentication.VC-CB provides scalability and flexibility, ideal for multi-party authenticationandsecureprinting.

FUTURE DIRECTIONS

Futureresearchcouldexplorehybridapproachescombining the strengths of both methods and investigate new

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 12 Issue: 06 | Jun 2025 www.irjet.net p-ISSN: 2395-0072

applicationsinemergingfieldssuchasIoTandblockchain and developing efficient algorithms for large-scale VC applications.

ACKNOWLEDGEMENT

The authors acknowledge the support received from the MAHAJYOTI, Nagpur and School of Technology, Swami Ramanand Teerth Marathwada University, Nanded, SubCampus,Peth,Laturforconductingthisresearch.

REFERENCES

[1] ShruthiHR,RanjanKumarHSandPrasannaKumarH R,“AVisualSecretSharingTechniqueforSecure And FastTransmissionofImage,”InternationalJournalof InnovativeResearchinComputerandCommunication Engineering,Vol.2,Issue4,April2014

[2] S. Shivani, “Multi secret sharing with unexpande meaningfulshares,”Multimed.ToolsAppl.,vol.77,no. 5,pp.6287-6310,Mar.2018,doi:10.1007/s11042-017453

[3] M.NaorandA.Shamir,“Visualcryptography”, in Proc. Adv.Cryptology-Eurocrypt,vol.950.1995,pp.1–12.

[4] O. Kafri and E. Keren, “Encryption of pictures and shapesbyrandomgrids,”Opt.Lett.,vol.12, no.6,pp. 377–379,Jun.1987.

[5] S.J.Shyu,“Imageencryptionbyrandomgrids,”Pattern Recognition.,vol.40,no.3,pp.1014–1031,2007.

[6] Y.F.Chen,Y.K.Chan,C.C.Huang,M.H.Tsai,Y.P.Chu, “Amultiple-levelvisualsecretsharingschemewithout imagesizeexpansion”,Inform.Sci.177(21) (2007) 4696–4710.

[7] T.H.ChenandK.H.Tsao,“Thresholdvisualsecret sharingbyrandomgrids,”J.Syst.Softw.,vol.84,no.7, pp.1197–1208,2011.

[8] Y.Hou,S.WeiandC.Lin,“Random-Grid-BasedVisual CryptographySchemes,”IEEETransactiononCircuits andSystemsforVideoTechnology,vol.24,No.5,May 2014

[9] ChandramathiS.,RameshKumarR.,SureshR.and HarishS.,“Anoverviewofvisualcryptography,”Int.J. Comput.Intell.Tech.,vol.1,no.1,pp.32-37,2010.

[10] G.Ateniese,C.Blundo,A.DeSantis,andD.R.Stinson, “Visualcryptographyforgeneralaccessstructures,” Inf.Comput.,vol.129,no.2,pp.86-106,Sep.1996, doi:10.1006/inco.1996.0070.

[11] G.Ateniese,C.Blundo,A.DeSantis,andD.R.Stinson, “Extendedcapabilitiesforvisualcryptography,”

TheoreticalComputerScience,vol.250,no.1-2,pp. 143-161,Jan.2001,doi:10.1016/S03043975(00)00220-5.

[12] T.H.ChenandK.H.Tsao,“Thresholdvisualsecret sharingbyrandomgrids,”J.Syst.Softw.,vol.84,no.7, pp.1197–1208,2011,doi:10.1016/j.jss.2011.02.017.

[13] Y.Hou,S.Wei,andC.Lin,“Random-Grid-BasedVisual Cryptography Schemes,” IEEE Trans. Circuits Syst. VideoTechnol.,vol.24,no.5,pp.733-744,May2014, Doi: 10.1109/TCSVT.2013.2273735.

[14] R.R.M.KumarandS.P.Raj,“Asurveyonvisual cryptographyschemes:Techniquesandapplications,” JournalofInformationSecurity,vol.5,no.2,pp.87101,2014,doi:10.4236/jis.2014.52010.

[15] Y.F.Chen,Y.K.Chan,C.C.Huang,M.H.Tsai,andY.P. Chu,“Amultiple-levelvisualsecretsharingscheme withoutimagesizeexpansion,”Inform.Sci.,vol.177, no. 21, pp. 4696–4710, 2007, doi: 10.1016/j.ins.2007.04.018.

[16] T.H.ChenandK.H.Tsao,“Thresholdvisualsecret sharingbyrandomgrids,”J.Syst.Softw.,vol.84,no.7, pp. 1197–1208, Jul. 2011, doi: 10.1016/j.jss.2011.02.017.

[17] S.J.Shyu,“Imageencryptionbyrandomgrids,”Pattern Recognition,vol.40,no.3,pp.1014–1031,2007,doi: 10.1016/j.patcog.2006.08.003.

[18] M.K.Lee,J.J.Kim,andS.H.Kim,“Asecureand FlexibleVisualcryptographyschemeusingcodebook forprotectingimagesecrets,”JournalofVisual CommunicationandImageRepresentation,vol.29, pp. 166-175, Jul. 2015, doi: 10.1016/j.jvcir.2015.04.010.

[19] G.Deepa,“TheComparativeStudyonVisual CryptographyandRandomGridCryptography.”IOSR JournalofComputerEngineering(IOSR-JCE),Volume 12,Issue,PP04-14,2May-Jun.2013.

[20] H.K.Lee,T.T.Lee,andS.W.Lee,“Efficientandflexible visualcryptographyschemewithsecretimage recovery,”JournalofComputerScienceand Technology,vol.25,no.4,pp.761–773,Jul.2010,doi: 10.1007/s11390-010-0976-2.