Mini Project onGENERATION OF MEANINGFUL SHARES IN VISUAL

CRYPTOGRAPHIC SYSTEMS

Presented By

Uttari Chakraborty (200911117) Debarko De (200915026)

Pratarshi Saha (200911075)

Under the supervision of

Mr. Sandeep Gurung(Associate Professor, CSE Dept., SMIT)

DEPT. OF COMPUTER SCIENCE AND ENGINEERINGSIKKIM MANIPAL INSTITUTE OF TECHNOLOGY

MAJITAR SIKKIM, 737136.Dated : 07/12/2012

CONTENTS

1.Introduction

2.Design Strategy

3.Test Plan

4.Implementation Details

5.Results and Discussions

6.Summary and Conclusion

INTRODUCTIONVisual Cryptography combines perfect ciphers

and graphics

Input = Binary ImageOutput = 2 (or more) shares

Shares can be printed on transparent papers

Overlapping of shares reveals the secret

No computer participation is required

Can be decoded directly by the HVS

AIM To implement Generation of meaningful shares in Visual

Cryptographic Systems

Problem Definition•To share secrets without revealing the

message

•A third party cannot decrypt the message if he has less than the required number of shares

•The main areas of problem are-The contrasts of the output images are lower

than the original secret message

To make the shares completely unsuspicious

Analysis of the Problem•To extract meaningful shares out of

various kinds of images•The schemes developed have a

disadvantage of contrast loss •Also the randomness of the shares make

them obvious•Hence, we have to develop a method to

conceal the shares inside meaningful images

Solution Strategy•Partial images are generated of the secret

message by breaking the image into random parts

•Distribute these parts among random images which upon overlapping reveal the secret message

•These partial images are embedded in meaningful shares which make it difficult to detect any hidden image

Literature SurveyVisual Cryptography and Shares

Visual Cryptography by Moni Naor and Adi Shamir, LNCS, Springer, Heidelberg (1994)

Different schemes for generation of sharesA Comprehensive Study of Visual Cryptography by Jonathan Weir and WeiQi Yan (Queen’s University Belfast, Belfast, BT7 1NN, UK)

Design Strategy

Fig:- Steps involved in accomplishing the project

TEST PLAN• Should work for any kind of image-

The algorithms used in this project work for any type, dimension or color model of image.

• Minimum contrast degradation-The PSNR of the algorithms is calculated to compare the unwanted contrast loss.

• Randomness in partial imagesAND operation is performed on the shares to check for randomness.

IMPLEMENTATION DETAILS(2,2) SCHEME

WHITE SHARES: { } BLACK SHARES: { }• Each pixel is divided into 2 sub pixels• Thereby distorts the aspect ratio of the image• This scheme is seldom used

(2,4) SCHEME

Problems

Contrast Ratio gets deviated from the original image and is almost half of that of the original image.

PROPOSED SOLUTIONHALFTONING

Half toning is a method for creating the illusion of Continuous tone output in Binary Mode.

•Error diffusion is a type of halftoning in which Quantization Residual is distributed to neighboring Pixels that have not yet been processed.

•Error diffusion using FLYOD-STEINBERG ALGORITHM

Initialize w(x,y) to Image(x,y) If w(x,y)

> threshold

Print 0

Print 1

NoYes

Compute Errorw(x,y)-max

Forward Distribute Error

Halftone Results

Original Gray Scale Image

Halftone ImageInput Image Output ImageHalftone

d

Comparison of (2,4) Scheme Output with and without Halftoning

Colour Visual Cryptography

Input ImageOuutput Image

RESULTS AND DISCUSSIONS

COMPARISON OF ALGORITHMS

PSNR Ratio is used to find out the relative execution time of different algorithms. The formaula to calculate the PSNR is as given below:

Type of Image PSNR Value

Black & White Image without Halftoning

6.5715 dB

Black & White Image with Halftoning

7.3651 dB

Colour CMYK Images 7.4337 dB

SUMMARY AND CONCLUSION• Thus, in this project we have implemented a new

type of cryptographic scheme which can decode concealed images without any computation.

• We have generated random parts of various types of images and embedded them in meaningful shares

• After generating shares using the cover image, we have overlapped those shares to get the decrypted output.

REFERENCES•Visual Cryptography by Moni Naor and

Adi Shamir, LNCS, Springer, Heidelberg (1994)

•A Comprehensive Study of Visual Cryptography by Jonathan Weir and WeiQi Yan (Queen’s University Belfast, Belfast, BT7 1NN, UK)

GANTT CHART

Project Gantt Chart

7 Aug 2012

11 Aug 2012

22 Aug 2012

15 Nov 2012

Problem Definition & Solution Formulatio

n Literature Survey

IMPLEMENTATION

Verification And Validation

Dcumentation

20 Oct 2012

Thank You