590

Short Paper Proc. of Int. Colloquiums on Computer Electronics Electrical Mechanical and Civil 2011

Performance Analysis of DWT and IWT on Non LSB Steganography H S Manjunatha Reddy1, Khasim T2, K B Raja2, 1

Global Academy of Technology/Department of ECE, Bangalore, India manjunathareddyhs@rediffmail.com 2 University Visvesvaraya College of Engineering/Department of ECE, Bangalore University, Bangalore, India, raja_kb@yahoo.com after a suitable transform domain leads to spatial domain and frequency domain steganography.

Abstract- Steganography is the art and science of writing hidden information for secure Communication. In this paper we propose Performance Analysis of DWT and IWT on Non LSB Steganography (PADIS). The cover image is segmented into 4*4 cells and DWT/IWT is applied on each cell. The 2*2 cell of HH band of DWT/IWT are considered and manipulated with payload bit pairs using identity matrix to generate stego image.

The key is used to extract payload bit pairs at the destination.

It is observed that the PSNR values are better in the case of IWT compare to DWT for all image formats. The PSNR values are high in the case of raw images compared to formatted images.

II. RELATED WORKS A J Mozo et al., [4] proposed Video Steganography using Flash Video. The software implementation in the field of video steganography is presented. A C++ program was implemented which could embed any type of the data in the Flash Video and could extract the same hidden data. Neha agarwal and Marios Savvides [5] proposed biometric Data Hiding: A 3 Factor Authentication Approach to Verify Identity with a Single Image using steganography, Encryption and Matching.

Hiding biometric data in DCT coefficients of cover image which can be exploited for hiding any kind of information image and biometric data like fingerprints and iris codes. MinWen Chao et al., [6] proposed A High capacity 3D Steganography algorithm. The scheme is based on a novel multilayered embedding scheme to hide secret messages in the vertices of 3D polygon models. Mamta Juneja and Parvinder Singh Sandhu [7] proposed a Robust Image Steganography Technique based on LSB insertion and encryption. Wu Ling and Wang Li-Na [8] proposed a Steganography algorithm based on Cyclic Redundancy Check (CRC). This method does not require matrix multiplication hence embedding/extracting codes perform better than matrices embedding schemes. Shiva Kumar et al., [9] proposed Cover Image Modulation Approach to Steganography in which cover image is segmented into 2*2 matrices and adjacent pixel intensity differences are computed to the threshold value to consider matrix for payload embedding.

Index Terms- Steganography, Cover Image, Payload, Stego image, DWT, IWT

I. INTRODUCTION The use of modern communication demands security in computer networks because of an increased use of internet.

The confidentiality and integrity of the data being transferred is to be protected against hackers. Security of information is one of the most important factors of information technology and communication. The steganography is a technique of hiding the secret messages in some medium such as text files, images, audio file, video files etc, so that it cannot be detected and doesn’t have any suspicion about the transmission of hidden or secret information. The word steganography [1] is derived from the Greek word steganos which means covered and graphy means writing. Steganography [2] is a method of hiding data in the Least Significant Bits (LSB) of a cover file.

The steganography technique has to satisfy three requirements viz., (i) Capability is the amount of payload embedded into the cover image. (ii) Imperceptibility is the stego image must be similar to the cover image. (iii) Security is the hidden payload in the cover image should not be disturbed by the noise in the communication channel and also not to reveal to hackers. There must be tradeoff between three parameters to have better steganographic technique.

Choosing the hiding medium as the criteria, the steganographic techniques [3] are classified as (i) Text- or linguistic based Steganography uses written natural language to conceal secret information. (ii) Audio Steganography embeds the message into a cover audio file as noise at a frequency out of human hearing range by LSB manipulation, phase coding and echo hiding. (iii) Image Steganography hide the secret message in the images which is nearly impossible to differentiate by human eyes. Based on the data hidden in the pixels directly or in the coefficients obtained Š 2011 ACEEE DOI: 02.CEMC.2011.01. 590

III. PROBLEM STATEMENT The payload is embedded into the cover image using nonLSB technique. Cover Image is segmented and DWT/IWT is applied to generate 2*2 cells. The payload bit pairs are manipulated with 2*2 cells of DWT/IWT to generate stego object.

IV. PROPOSED MODEL In this section definitions of evaluation parameters and proposed embedding and retrieval system are described.

A. Evaluation Parameters (i) Mean Square Error (MSE): It is defined as the square of error between cover image and stego image. The distortion 42

Short Paper Proc. of Int. Colloquiums on Computer Electronics Electrical Mechanical and Civil 2011 in the image can be measured using MSE and is calculated using Equation (1).

(d)If the bit pair is (1, 0) then first row is interchanged. (vi) Multiplier: Multiply IM with 2x2 cover image which results stego matrix, IM X = K (vii) Key: It is used to extract the payload at the estination so that first bit of the payload bit pair is taken as reference in order to avoid unintended recipient. (viii) Inverse Wavelet Transform: The 2*2 matrices are obtained from the Multiplier to form stego DWT/IWT matrices. Inverse DWT/IWT is applied to generate stego image.

Where X(i,j): Cover image pixel intensity value Y(i,j): Stego image pixel intensity value N: Size of an Image. (ii) Peak Signal to Noise Ratio (PSNR): It is the measure of quality of the image by comparing the cover image with the stego image, i.e., it gives the statistical difference between the cover image and stego image and it is calculated using Equation (2).

PSNR  10 log10

255

C. Payload Extraction The payload is extracted from the stego image at the destination by adapting reverse process of embedding.

2

dB (2) MSE (iii) Capacity: It is the size of the secret message embedded in a cover image that can be modified without deteriorating the integrity of the cover image B. Proposed PADIS Embedding Technique The proposed PADIS embedding algorithm is shown in Figure 1. The cover image is segmented into 4*4 cells and DWT/ IWT is applied on each cell to form 2*2 sub band matrices.

The embedding process is performed on Horizontal band of DWT/IWT. (i) Cover Image (CI): It carries original secret information.

The different formats like raw images, JPEG, BMP, TIFF and PNG are considered to test the algorithm with different sizes. (ii) Payload (PL): It is the secret information that can be embedded into the cover image to generate stego image. (iii) Segmentation: The cover image of size is decomposed into matrix of size 4*4 to enhance the security to the payload. (iv) Wavelet Transform (WT): The two dimensional DWT/ IWT applied on each 4*4 matrix of cover image to convert into transform domain of size 2*2. The four sub bands viz., LL, HL, LH and HH are obtained for DWT and IWT. For embedding process, the HH bands of DWT/IWT are considered. The HH coefficients of DWT are quantized to get integer values. (v) Modification: All the columns of payload pixel values are converted into single column. These pixels are converted in to binary. For hiding data, two payload bits (b1, b2) are considered at a time. In this technique, the MSB of all pixels are considered first for hiding. We can choose payload bits in any manner.

Figure1. PADIS embedding system

V. EMBEDDING ALGORITHMS The embedding of payload using PADIS algorithm is given in table 1.

T ABLE I. PADIS EMBEDDING ALGORITHM

1 0  Consider the identity matrix I =  0 1  Modification matrix IM is obtained from I using the conditions given below; (a) If the both bit pair of payload are (0, 0) then there is no change in ‘I’. (b) If both bit pair are (1, 1) then both rows of I are interchanged. (c)If the bit pair is (0, 1) then second row is interchanged, © 2011 ACEEE DOI: 02.CEMC.2011.01. 590

43

Short Paper Proc. of Int. Colloquiums on Computer Electronics Electrical Mechanical and Civil 2011 VI. RESULTS AND DISCUSSION

VII. CONCLUSIONS AND FUTURE WORKS

Cover image of different sizes and formats viz., JPEG, BMP, PNG and Raw images are considered for performance analysis.

Steganography performance is analysis using PSNR. The few CI images Garden, Football fans, House and Water house and few PL images are Cups, Lion, Sony and Rose are considered. Table 2 and 3 gives the PSNR between cover image and stego image for different image formats. It is observed that the PSNR is better using IWT in comparison with DWT for all image formats. The PSNR values are high in the case of raw images compared to formatted images. Since raw images has more redundant information.

Steganography is used for secure communication. In this paper PADIS algorithm is proposed. The cover image is segmented into 4*4 matrices. The DWT/IWT is applied on each cell to derive sub bands. The Payload bit pairs are manipulated with 2*2 bands of DWT /IWT to derive stego image. It is observed that PSNR value is better in the case of IWT compare to DWT. In future the algorithm can be tested with some more wavelet transform domain techniques to improve the performance.

T ABLE II. PSNR BETWEEN C OVER IMAGE AND STEGO IMAGE FOR RAW IMAGES AND JPEG FORMATS.

[1] Daniela Stanescu, Valentin Stangaciu, Loana Ghergulescu and Mircea Stratulat, “Steganography on Embedded Devices,” IEEE International Symposium on Applied Computational Intelligence and Informatics, pp.313-317, 2009.

[2] Abbas Cheddad, Joan Condell, Kevin Curran and Paul Mc Kevitt, “Enhancing Steganography in Digital Images,” IEEE Conference on Computer and Robot Vision, Canada pp.326-332, 2008.

[3] Kotaro Yamamoto and Munetoshi Iwakiri, “A Standard MIDI File Steganography Based on Fluctuation of Duration,” International Conference on Availability, Reliability and Security, pp.774-779, 2009.

[4] A J Mozo, M E Obien, C J Rigor, D F Rayel, K Chua and G Tangonan, “Video Steganography using Flash Video,” International Instrumentation and Measurement Technology Conference, pp.822827, 2009.

[5] Neha Agrawal and Marios Savvides, “Biometric Data Hiding: A 3 Factor Authentication Approach to Verify Identity with a Single Image Using Steganography, Encryption and Matching,” IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, pp.85-92, 2009.

[6] Min-Wen Chao, Chao-hung Lin, Cheng-Wei Yu and Tong-Yee Lee, “A High Capacity 3D Steganography Algorithm,” IEEE Transactions on Visualization and Computer Graphics , pp.274284, 2009.

[7] Mamta Juneja and Parvinder Singh Sandhu, “Designing of Robust Image Steganography Technique Based on LSB Insertion and Encryption,” IEEE International Conference on Advances in Recent Technologies in Communication and Computing, pp.302305, 2009.

[8] Wu Ling and Wang Li Na, “A Novel Steganography Algorithm Based on CRC,” IEEE International Conference on Multimedia Information Networking and Security, pp.122-125, 2009.

[9] K B Shiva Kumar, T G Manoj, K B Raja, Sabyasachi Pattnaik and R K Chhoparay, “Cover Image Modulation Approach to Steganography,” International Conference on Computer Modeling and Simulation, pp.V2-43-V2-47, 2011.

PSNR

BETWEEN

T ABLE III. C OVER IMAGE AND STEGO IMAGE FOR BMP AND PNG

REFERENCES

IMAGE

FORMATS

© 2011 ACEEE DOI: 02.CEMC.2011.01. 590

44