HIDING SECRET IMAGES AND MESSAGES USING IMAGE STEGANOGRAPHY Ram Prasanth_Hiding... · 1 . HIDING SECRET IMAGES AND MESSAGES USING IMAGE STEGANOGRAPHY By . Sai Ram Prasanth (EE10B097)

1

HIDING SECRET IMAGES AND MESSAGES USING IMAGE STEGANOGRAPHY By

Sai Ram Prasanth

(EE10B097)

Thesis of this project will be Submitted to the faculty of IIT Madras as a part

of my academic curriculum in pursuing the degree for B.tech and M.tech

In Electrical Engineering

PROJECT WORK AT

Institute of Development and Research in Banking Technology (IDRBT)

Road No. 1, Castle Hills, Masab Tank, Hyderabad-500057

Under the esteemed guidance of:

(Internal Guide) (ExternalGuide) Dr.Arun Pachai Kannu Dr. Rajarshi Pal Professor, Electrical Engg. Dept. AssistantProfessor IIT Madras IDRBT, Hyderabad

Department of Electrical Engineering Indian Institute of technology-Madras, Chennai 600 036

2

Institute of Development and Research in Banking Technology (IDRBT) Road No. 1, Castle Hills, Masab Tank, Hyderabad-500057

HIDING SECRET IMAGES AND MESSAGES USING IMAGE STEGANOGRAPHY Applications in Banking technology

CERTIFICATE

This is to certify that Mr Sai Ram Prasanth, pursuing Btech in Electrical Engineering at IIT Madras has undertaken a project as a Research intern at IDRBT, Hyderabad from May 20, 2013 to July 20,2013. He was assigned the project “HIDING SECRET IMAGES AND MESSAGES USING IMAGE STEGANOGRAPHY” under my guidance.

I wish him all the best for all his endeavours.

Dr. Rajarshi Pal (Project Guide) Assistant Professor IDRBT, Hyderabad

3

Acknowledgment

I express my deep sense of gratitude to my Guide Dr. Rajarshi Pal, Assistant Professor, IDRBT for giving me an opportunity to do this project in the Institute for development and research in Banking Technology and providing all the support and guidance needed which made me complete the project on time. I am thankful to Mr. Naveen Chandra, Research Associate, IDRBT who took keen interest in my project work and guided me all along till the completion of my project work by providing me all the necessary information. I am also thankful to IIT Madras for giving me this golden opportunity to work in a high-end research institute like IDRBT. Sai Ram Prasanth 3rd Year B.tech Electrical Engineering Dept. IIT Madras.

4

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.1 Applications in banking technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.2 Scope and objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 2 High capacity Image steganography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.1 Experimental results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

3 Secured image digital steganography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10

3.1 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 4 High capacity and security steganography using discrete wavelet transform. . . . . . . . . . . . . . . . . . . 13

4.1 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

6 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

5

[1]Introduction

Steganography: Steganography is an art of conveying secret messages and secret images through cover images in a secret way that only the receiver knows the

existence of a message.

Steganography Vs Cryptography:

•Cryptography is the practice of ‘scrambling’ messages so that even if detected, they are very difficult to decipher. The purpose of Steganography is to conceal the message such that the very existence of the hidden message is concealed. •Steganography provides another layer of protection on the secret message, which will be embedded in another media such that the transmitted data is meaningful and innocuous to everyone.

Fundamental requirements for a Steganography

•Imperceptibility: Means that the embedded messages should not be discernible to the human eye. •Embedding Capacity: Means the capacity of embedding the secret image. •Security: Means that the Stego image should be fool proof and robust.

[1.1]Applications in Banking technology

Due to unavoidable hacking of the databases on the internet, it is always quite difficult to trust the information on the internet To solve this problem of authentication, there are many algorithms based on image processing, improved steganography and visual cryptography These method conceal the secret messages within some other ordinary media (i.e. images, music and

6

video files) so that it cannot be observed. The main objective of steganography is to communicate securely in such a way that the true message is not visible to the observer. That is unwanted parties should not be able to distinguish in any sense between cover-image (image not containing any secret message) and stego-image (modified cover-image that containing secret message). Thus the stego-image should not deviate much from original cover-image. Today steganography is mostly used on computers with digital data being the carriers and networks being the high speed delivery channels. It is also used in banking technology for communicating confidential data between two persons.

[1.2]Scope and objecticve • To test the usefulness of several image steganographic models in banking. • To achieve this purpose we have implemented there steganographic

models on 1. High capacity image steganography[1] 2. Secured Digital Image Steganography[2] 3. Secured Digital Image Steganography[3]

7

[2]High capacity image steganography In this paper[1], a high capacity image steganographic model based on variable size LSB insertion is proposed, the embedding capacity will be over 50% of the cover-image size, and the image fidelity is better than four-LSBs insertion. The most important advantage is that no artifacts appear in the stego-image with high embedding capacity. Proposed Model:

Embedding module :

CE component: The HVS is insensitive to the noise component and the psycho-visually redundant component (bit planes from 5-8)in an image, thus these components can be used to embed messages.

8

So the embedding capacity K(x, y) of each pixel can be computed by the following expression: K(x,y) = min{max[Kn(x,y), 4], U(x,y)} MER component: It improves the image fidelity by adjusting the N+1 bit for an N bit embedding, so that the grey value doesn’t vary much with the original cover image. Let f(x, y) be the original grey scale g(x, y) be the gray scale obtained by embedding k LSBs directly, and g‘(x, y) be the grey scale obtained by changing the value of the (k+ 1)th LSB of g(,x, y). The minimum-error grey scale must be g(x, y) or g’(x, y). Let e(x, y) be the error between f(x, y) and g(x, y). and e’(x, y) be the error between f(x, y) and g‘(x, y). If e(x, y) < e’(x, y), then g(x, y) will be used to replace f(x, y); otherwise g’(x, y) is selected. Using this method the maximum embedding error can be restricted to 2(𝑘−1). lGSC component: Embedding too many bits in the smooth area of an image will cause the false contouring. In order to fix this IGSC component is imposed by distributing the extra grey value equally among its neighbours at every smooth pixel. In our IGSC component the embedding error is evenly spread to the bottom and right neighbouring pixels Let e(x, y) denote the embedding error of pixel ‘p’ at co-ordinates (x, y), these four bottom-right neighbouring gray scales are then modified according to the following expressions:

9

[2.1]Experimental Results

The average embedding capacity of each pixel is 4.06 bits per pixel,

After CE component: RMS: 6.62 PSNR: 3I.74 dB (Zero BER while extracting)

After MER component: RMS: 5.73 PSNR: 32.97dB (Zero BER while extracting)

After IGSC component: RMS: 5.17 PSNR: 34.23dB (2.32% BER while extracting)

Method Capacity RMS PSNR Artifacts

4 LSBs insertion

4 bits/pixel 6.61 31.71 Yes

Proposed method

4.06 bits/pixel

6.02 32.57 No

Since there is significant bit error rate associated with IGSC component while retrieving, it won’t be precise in using for text and numeric data. But since the values of PSNR are high we can use it for hiding the images, by exploiting the human visual perceptibility.

10

[3]Secured Digital Image Steganography This paper[2] presents a new method for data hiding into the discrete wavelet coefficients of the cover image in order to maximize the hiding capacity overcome the drawback. In this method, they used two processes. The first one is encoding and second one is decoding process. Encoding: In encoding, they will apply Arnold transform with key on secret image and get the scrambled secret image. This process gives the more security and robustness to our algorithm. Apply DWT on the cover image and scrambled secret image in order to increase the security level. The alpha blending matrix is obtained, by the addition of wavelet coefficients of respective sub-bands of cover image and scrambled secret image. Alpha factor is increasing the embedding strength factor. Once the Alpha blending operation is done, then apply the Inverse discrete wavelet transform (IDWT) and get the stego image. Decoding: The decoding process is actually the reverse process of the embedding model.

Encoding Process

11

Decoding Process [3.1]Experimental results

DWT (discrete wavelet transform)

The wavelet-based transform uses a 1-D sub band decomposition process in which a 1-D set of sample is converted into the low pass sub band (Li) and high-pass sub band (Hi). Where i represents the level of decomposition.

The low-pass sub band represents a down sampled low-resolution version of the original image. The high-pass sub band represents residual information of the original image.

In 2-D sub band decomposition, the entire process is carried out by executing 1-D sub band decomposition twice, first in one direction (horizontal), then in the orthogonal (vertical) direction

Alpha Blending: alpha blending matrix is obtained, by the addition of wavelet coefficients of respective sub-bands of cover image and scrambled secret image.

12

Cover image

Secret Image

MSE PSNR NCC AD SC MD NAE

Deer.jpg 316X 380

Baby.jpg 458 X 500

0.7542 49.320 0.9943 0.507 1.011 2.860 0.0065

Deer.jpg 316X 380

Flwr1.jpg 300 X450

1.0440 47.943 0.9925 0.743 1.015 3.148 0.0078

Deer.jpg 316X 380

Flwr2.jpg 564 X395

0.9249 48.470 0.9949 0.430 1.010 3.176 0.0067

coconut.jpg 316X 380

Flwr2.jpg 564 X395

0.9538 48.336 1.0058 0.640 0.998 3.554 0.0227

MSE: Mean square error PSNR: Peak Signal to Noise Ratio NCC: Normalized cross correlation AD: Average difference SC: Structural content MD: Maximum difference NAE: Normalized absolute error

Using wavelet transform technique a secret image of size 2N*2N can be encrypted in a cover image of size N*N. It can be further increased by increasing the level of DWT transformation. Since high capacity embedding is possible with reasonable PSNR values, this method can also be applied for bulk. In addition, this method gives more capacity and high security to transfer images in communication field. Experimental results show that this method gets stego-image with perceptual invisibility, high security and certain robustness.

13

[4] HIGH CAPACITY AND SECURITY STEGANOGRAPHY USING DISCRETE WAVELET TRANSFORM

In this paper[3] High Capacity and Security Steganography using Discrete wavelet transform (HCSSD) is proposed. The wavelet coefficients of both the cover and payload are fused into single image using embedding strength parameters alpha and beta. The cover and payload are preprocessed to reduce the pixel range to ensure the payload is recovered accurately at the destination. It is observed that the capacity and security is increased with acceptable PSNR in the proposed algorithm compared to the existing algorithms. Wavelet Transform: Wavelet transform is used to convert a spatial domain into frequency domain. The use of wavelet in image stenographic model lies in the fact that the wavelet transform clearly separates the high frequency and low frequency information on a pixel by pixel basis. Discrete Wavelet Transform (DWT) is preferred over Discrete Cosine Transforms (DCT) because image in low frequency at various levels can offer corresponding resolution needed. A one dimensional DWT is a repeated filter bank algorithm, and the input is convolved with high pass filter and a low pass filter. The result of latter convolution is smoothed version of the input, while the high frequency part is captured by the first convolution. The reconstruction involves a convolution with the synthesis filter and the results of this convolution are added. In two dimensional transform, first apply one step of the one dimensional transform to all rows and then repeat to all columns. This decomposition results into four classes or band coefficients.

Proposed method ALGORITHM OF DATA EMBEDDING Input: Cover Image c and Payload image p. Output: Stego image s. 1. Normalize c and p, so that the wavelet coefficient varies between 0.0 and 1.0.

14

2. Preprocessing on c and p 3. Transform c and p into 2 levels of decomposition using Haar Wavelet. 4. Apply 2 levels DWT on the approximate band of the payload obtained. 5. Encrypt the DWT coefficients obtained. 6. Wavelet fusion of DWT coefficients of c and p. 7. Inverse transform of all the subbands of the fused image. 8. Denormalize the Fused image. 9. Stego image s is generated. ALGORITHM OF DATA EXTRACTION Input: Stego Image s. Output: Payload p. 1. Normalize Stego Image s. 2. Transform s in to 2 levels of wavelet decomposition. 3. Subtract DWT coefficients of c from DWT coefficients of s to get DWT coefficients of only p. 4. Decrypt the DWT coefficients of p obtained. 5. Apply IWT of all the sub bands of p. 6. Apply IWT of payload obtained with respect to approximate band. 7. Denormalize the resultant of step 6. 8. Payload Image is obtained p.

HCSSD Encoder

15

HCSSD Decoder

[4.1] Experimental Results

Images Type Size MSE PSNR Entropy

Lady Flower

JPEG JPEG

346×396 240×240

0.17 55.6 0.00019

Aero plane Bank Text

TIFF PNG

400×300 810×400

2.76 43.7 0.0000

Player Astronauts

JPEG PNG

400×300 200×200

0.9 48.1 0.0004

Cow Boys Dog

JPEG TIFF

186×100 436×600

0.17 55.58 0.0000

16

[5]Conclusion

The first technique is basically designed for high capacity encryption, the fidelity of the stego image is increased by adding MER and IGSC components.

• Retrieval with zero BER is possible with MER and CE components so they can be used for text messages and images.

• IGSC can’t be recovered exactly so it is limited in transmitting images.

The second model mainly concentrates on Robustness and high capacity encryption using wavelet transform technique

• Retrieval with Zero BER is possible with many sized images, so this can be used for both image and text message transmission

The third model concentrates on both High Capacity and Security Steganography

• It is observed that the capacity and security is increased with acceptable PSNR in the proposed algorithm compared to the existing algorithms.

17

[6]References

[1] Y. K. Lee. and L. H.Chen. High capacity image steganographic model.IEE Proc.-Vision, Image and Signal Processing, 147:288–294, 2000.

[2] Prabakaran. G and Bhavani. R, "A Modified Secure Digital Image Steganography Based on Discrete Wavelet Transform", International Conference on Computing, Electronics and Electrical Technologies, pp. 1096-1100, 2012. [3] H S Manjunatha Reddy and K B Raja, "High Capacity and Security Steganography using Discrete Wavelet Transform", International Journal of Computer Science and Security (IJCSS), Volume (3): Issue (6), pp. 462-472.