Comparative Study on Watermarking & Image Encryption for
Transcription
Comparative Study on Watermarking & Image Encryption for
INTERNATIONAL JOURNAL FOR TRENDS IN ENGINEERING & TECHNOLOGY VOLUME 5 ISSUE 1 – MAY 2015 - ISSN: 2349 - 9303 Comparative Study on Watermarking & Image Encryption for Secure Communication Dr. 2S. Poornachandra2 Ajna Madanan1 Anna University, ECE dean.iqac.snsce@gmail.com 1 AnnaUniversity, ECE, ajna.madanan@gmail.com Abstract—Over the past decades, research in security has concentrated on the development of algorithms and protocols for authentication, encryption and integrity of data. Despite tremendous advances, several security problems still afflict system’s. In this android app watermarking and encryption is being applied on images and data. Because of the human visual system’s low sensitivity to small changes and the high flexibility of digital media, anyone can easily make small changes in digital data with low perceptibility. Here watermarking and encryption are being performed in wavelet domain. Here in watermarking, the coefficients of watermarks are being embedded with the coefficients of the original image. Encryption is being done in wavelet domain so that the probability of an intruder trying to access the contents is very much minimized. Thus, this model provides a high level of security. Index Terms— Security, wavelets, watermarking, encryption, intruder. —————————— —————————— 1 INTRODUCTION E xplosive growth in multimedia technology and its applications has escalated the necessity to build secure methods for legal distribution of the digital content. Presently need for protection against unauthorized copying and distribution has increased. Thus, there aroused apprehensions over copyright protection of digital contents. The solution to this problem is digital watermarking, which is the most common and possibly the strongest technique for protecting digital data. Cryptography ensures that the messages are secure from possible ―attacks‖, which are impersonation, eavesdropping etc. In image watermarking, information is embedded into cover media for proving ownership identification. Various watermarking techniques have been proposed by many authors in the last several years, which include spatial domain and transform domain watermarking. To provide security of the images in the multimedia environment, encryption plays an important role. Image encryption ensures confidential transmission and storage of image over internet. Image encryption has applications in multimedia systems, internet communication, telemedicine, medical imaging, and military communication to name a few. Watermarking is a method of data embedding and information hiding. In digital watermarking, patterns of bits are inserted into a digital image, video or audio file that helps in identifying the copyright information, i.e. author, rights etc. It is a concept closely related to steganography. Watermarking is considered as a special technique of steganography where one message is embedded in another and the two messages are related to each other in some way. Steganography and cryptographic methods are used together with wavelets to increase the security of the data while transmitting through networks. A combination of both encryption with steganography allows better private communication. The goal of steganography is to avoid drawing suspicion to the transmission of the secret message. In digital watermarking a low energy signal is imperceptibly embedded into another signal. The low-energy signal is known as watermark. The main signal in which the watermark is embedded is referred to as cover signal since it covers the watermark. The entity known as watermark key is used for embedding and detecting In this paper RSA algorithm is being used for encryption. The RSA encryption technique is being named after the inventors Ron Rivest, Adi Shamir, and Len Adleman in 1977. RSA is a public-key cryptography system. This algorithm is based on the difficulty of factorizing large numbers that have two and only two factors (prime numbers). It works on a public key system. Everyone would know the public key. Using public key a user can encrypt data, but cannot decrypt it. The person who possesses the private key can only decrypt it. Theoretically, it may be possible, but it is extremely difficult to generate the private key from the public key; this makes the RSA algorithm a very popular choice in data encryption. ———————————————— Ajna Madanan is currently pursuing masters degree program in electronics and communication engineering in Anna University, India, PH8220829679. E-mail: ajna.madanan@gmail.com V.J. Subashini is currently Associate Professor in Computer Applications in AnnaUniversity, India, PH-9444057965. E-mail: vjsubashini@yahoo.com Dr.S. Poornachandra is currently the dean /IQAC in electronics and communication engineering in Anna University, India, PH-8643844943. Email: dean.iqac.snsce@gmail.com watermark signal. 1 INTERNATIONAL JOURNAL FOR TRENDS IN ENGINEERING & TECHNOLOGY VOLUME 5 ISSUE 1 – MAY 2015 - ISSN: 2349 - 9303 information leakage. As the demand to protect the sensitive and valuable data from satellites has increased, a new method for satellite image security by combining DWT-DCT watermarking and AES encryption has been implemented [8]. 2 RELATED WORK Wavelet based image watermarking is gaining more popularity because of its resemblance to the human visual system. In the wavelet-based perceptual watermark masking proposed by Khalil Zebbiche and FouadKhelifi [1], the perceptual masking model takes into consideration local brightness, band sensitivity and texture masking to compute a weight for each DWT coefficient of the fingerprint image. Thus, the perceptual model consists of a spatial frequency sensitivity function and two masking components based on local brightness and texture masking. The watermarking scheme yields a better robustness (i.e. lower BER). The bit error ratio or bit error rate (BER) is the number of bit errors divided by the total number of transferring bits during a time interval. M. Barni [2] have developed an improved wavelet-based watermarking through pixel-wise masking. It is mainly based on masking the watermark according to characteristics of Human Visual System (HVS). The watermark is being adaptively added to the largest detail bands. The watermark weighing function is calculated as a simple product of data extracted from HVS model. The watermark is being detected by correlation. Victor [3] have developed an algorithm that relies upon adaptive image watermarking in high resolution sub-bands of DWT. N. Kaewkamnerd and K.R. Rao [4] developed a wavelet based image adaptive watermarking scheme. Embedding is performed at higher level sub-bands of wavelet transform, even though this may clearly change the image fidelity. Bo Chen and Hang Shen [5] have developed a new robust fragile double image watermarking algorithm using improved pixel-wise masking model and a new bit substitution based on pseudo-random sequence. The method embeds robust and fragile watermark into the insensitive part and sensitive part of wavelet coefficients making two watermarks non-interfering. To improve performance, discrete wavelet transform (DWT) has been combined with another equally powerful transform which is discrete cosine transform (DCT). The combined DWT-DCT watermarking algorithm’s imperceptibility was better than the performance of the DWT approach. For watermarking, the preferred color model must be HSV (Hue, Saturation and Value) rather than RGB because it is the most closely related color model with Human Visual System [9]. Salt and pepper noise of input color satellite image can be removed by using Modified decision based unsymmetrical trimmed median filter (MDBUTMF) algorithm. Prashant Shinde and Chetan Mohol [10] have proposed application for Android OS that permits to add an invisible or visible watermarking into an image of the current designations. 3 PROPOSED SYSTEM Here encryption is performed in wavelet domain so that the probability of an intruder trying to access the contents is very much minimized. Encryption is being provided after watermarking so that security level can be increased. 3.1 Transmission Section Peng Liu and Zhizhong Ding [6] proposed a blind image watermarking scheme based on wavelet tree quantization. Here largest two coefficients are selected as significant coefficients and the difference between them is taken as significant difference. Watermark bits are embedded by comparing the significant difference with an average significant difference value and maximum difference coefficients are quantized. Mohammad. A. B. Younes [7] introduced a block-based transformation algorithm based on the combination of image transformation and a well-known encryption and decryption algorithm called Blowfish. The original image is being divided into blocks that are then rearranged into a transformed image using a transformation algorithm, and then the transformed image is encrypted using the Blowfish algorithm. The results show that the correlation between image elements was significantly decreased by using this technique. The results also showed that increasing the number of blocks by using smaller block sizes resulted in a lower correlation and higher entropy. Fig. 1. Extracted watermark using Haar wavelet Here the original image and watermark are being converted to wavelet coefficients using Haar transform. The coefficients of the watermark are added with the coefficients of the original image to get the watermarked image. During the embedding process, the watermark coefficients are being embedded into the highest value wavelet coefficient of the original image. Here for the encryption, RSA algorithm is being used. While encrypting, keys are also used along with user-typed message which is being converted to wavelet coefficients. After encryption they are being transmitted. This is being shown in Fig. 1. Encryption is being provided after watermarking so that security level can be increased. The researches and applications of applying digital watermarking to Geo-information data are still very inadequate. Watermarking techniques that are developed for multimedia data cannot be directly applied to the satellite images because here the analytic integrity of the data, rather than the perceptual quality, is of primary importance. Also satellite images have higher requirements in content security. It desires not only watermarking for copyright protection, but also encryption during storage and transmission for preventing 2 INTERNATIONAL JOURNAL FOR TRENDS IN ENGINEERING & TECHNOLOGY VOLUME 5 ISSUE 1 – MAY 2015 - ISSN: 2349 - 9303 TABLE I 3.2 Receiver Section MSE AND PSNR VALUES USING HAAR WAVELET Haar Wavelet MSE 124.2513 PSNR 27.1878 Fig. 3. Extracted watermark using Haar wavelet In Table I and II, it can be seen that Haar wavelet and db 1 (Daubechies wavelet) produces the same PSNR and MSE values. From Fig. 3 and 4(a), it is seen that watermark extracted are as same as the original one. In Daubechies daughter wavelets, i.e. from DB 2 to DB 10 in Fig. 4, it is seen that spots are being introduced in the extracted watermark and also the left side is being distorted. Fig. 2. Extracted watermark using Haar wavelet TABLE II MSE AND PSNR VALUES USING DAUBECHIES WAVELET After the reception, the user-typed message would be converted to wavelet coefficients and then it would be used for decryption along with the keys. By applying inverse DWT, the image and watermark image would be returned. The watermark is removed at the receiving end with the help of the key and the original image. This process is called watermark extraction. In the present model, only threshold is being considered. This is being shown in Fig. 2. Daubechies Wavelet Daughter Wavelets MSE DB 1 124.2513 4 RESULT ANALYSIS Performance metrics used here are Mean Square Error and Peak Signal-to-Noise Ratio and visual comparison which are done through correlation. Mean Square Error (MSE) allows comparing the ―true‖ pixel values of the original image to the degraded image. MSE represents the average of the squares of the "errors" between actual image and noisy image. Errors are the amount by which the values of the original image differ from the degraded image. MSE can be found using the equation: PSNR 27.1878 DB 2 123.0029 27.2316 DB 3 121.3821 27.2893 DB 4 119.5124 27.3567 DB 5 117.6615 27.4245 DB 6 116.3122 27.4745 DB 7 114.9985 27.5239 DB 8 113.5129 27.5803 DB 9 111.6682 27.6515 DB 10 110.1306 27.7117 (1) The notations used are listed below. 𝑋 (𝑖, 𝑗): Original image, 𝑋′ (𝑖, 𝑗): Watermarked image, and Nt : Size of image (a) Peak Signal-to-Noise Ratio(PSNR) is an expression for finding the ratio between the maximum possible value (power) of a signal and the power of distorting noise that affects the quality of its representation. As many signals have a very wide dynamic range (ratio between the largest and smallest possible values of a changeable quantity), PSNR is frequently expressed in terms of the logarithmic decibel scale. PSNR can be found using the equation: (f) (2) Here Matlab is being used as the simulation software. On applying different wavelets such as Haar, Daubechies, Coiflet, Symlet on Fig. 3, various values of MSE and PSNR are as follows: (b) (g) (c) (d) (e) (h) (i) (j) Fig. 4. Extracted watermark using Daubechies daughter wavelet i.e. from DB 1 to DB 10. 3 INTERNATIONAL JOURNAL FOR TRENDS IN ENGINEERING & TECHNOLOGY VOLUME 5 ISSUE 1 – MAY 2015 - ISSN: 2349 - 9303 Table III more when compared to Symlet wavelet as seen in Fig. 6. Therefore, from the above analysis, it is being concluded that Haar wavelet and DB 1 of Daubechies wavelet produces better results by comparing extracted watermark with the original image. MSE and PSNR values using Symlet wavelet Symlet Wavelet MSE 123.0029 PSNR 27.2316 SYM 3 121.3821 27.2893 SYM 4 111.4139 27.6614 SYM 5 108.2783 27.7854 SYM 6 104.9880 27.9194 SYM 7 105.9100 27.8814 SYM 8 098.0600 28.2159 Daughter Wavelets SYM 2 (a) (b) (c) (f) (d) 5 CONCLUSION This paper introduces a model that provides a high level of security by using both watermarking and encryption in the wavelet domain. Due to this the probability of an intruder’s access to the contents is very much minimized. (e) (g) Fig. 5. Extracted watermark using Symlet daughter wavelet i.e. from SYM 2 to SYM 8. Similarly, in Symlet daughter wavelets, i.e. from SYM 2 to SYM 8, it is seen that spots are being introduced and its intensity increases in the increasing order of Symlet daughter wavelets. Also the left side is being distorted and it increases more while reaching SYM 8. This can be seen in Fig. 5. Table IV Fig. 7. Images a) Original Image b) Watermark Image c) Watermarked Image d) Extracted watermark Here watermarking has been implemented using both grayscale and color images and watermark was extracted successfully. Fig. 7 shows the original image, watermark, watermarked image and the extracted watermark. By comparing the original watermark image and the extracted watermark images, it is seen that Haar wavelet produced better results. MSE and PSNR values using Coiflet wavelet Coiflet Wavelet Daughter Wavelets MSE PSNR COIF 1 117.6857 27.4236 COIF 2 109.5243 27.7357 COIF 3 102.0552 28.0425 COIF 4 095.2216 28.3434 COIF 5 087.2121 28.7250 (a) (b) (c) (d) ACKNOWLEDGMENT I extend my heart-felt thanks to my Project Guide, Dr. S. Poornachandra, Dean, IQAC, for his remarkable guidance, encouragement, valuable suggestions and support for the completion of this Project. I take immense pleasure in expressing my humble note of gratitude to, Dr. Karthikeyan V.V., Head of Department for giving the permission to carry my project in the department. I am highly grateful to our PG Project Coordinator Dr. G.K.D. PrasannaVenkatesan, for his valuable suggestions throughout the course of this project. I record my indebtedness to our Principal Dr. N. Gunasekaran, for his guidance and sustained encouragement for the completion of the project. I also extend my thanks to Mrs. V.J. Subashini, Senior Assistant Professor, Jerusalem College of Engineering, Chennai, for her remarkable guidance, encouragement, valuable suggestions and support for the completion of this Project. I also extend my thanks to Faculty members, my parents, and friends for providing their guidance and moral support in successfully completing the Project. (e) Fig. 6. Extracted watermark using Coiflet daughter wavelet i.e. from COIF 1 to COIF 5 Similarly from Table III and IV it is seen that degradation of image quality increases as one goes down. Degradation in Coiflet is 4 INTERNATIONAL JOURNAL FOR TRENDS IN ENGINEERING & TECHNOLOGY VOLUME 5 ISSUE 1 – MAY 2015 - ISSN: 2349 - 9303 REFERENCES [1] Khalil Zebbiche and FouadKhelifi, ―Efficient wavelet-based perceptual watermark masking for robust fingerprint image watermarking‖, IET Image Process, Vol. 8, Iss. 1, pp. 23–32 doi: 10.1049/iet-ipr.2013.0055, 2014. [2] Barni M, Bartolini F, Piva, ―An Improved Wavelet Based Watermarking Through Pixelwise Masking‖, IEEE transactions on image processing, Vol. 10, 2001 pp.783-791. [3] Victor V., Guzman, Meana, ―Analysis of a Wavelet-based Watermarking Algorithm‖, IEEE Proceedings of the International Conference on Electronics, Communications and Computer, 2004, pp. 283-287. [4] N. Kaewkamnerd and K.R. Rao, ―Wavelet Based Image Adaptive Watermarking Scheme‖, IEEE Electronic Letters, Vol. 36,Feb. 2000, pp.312-313. [5] Bo Chen, Hong Shen, ―A New Robust-Fragile Double Image Watermarking Algorithm‖, Third IEEE International Conference on Multimedia and Ubiquitous Engineering,2009, pp. 153-157 [6] Peng Liu, Zhizhong Ding, ―A Blind Image Watermarking Scheme Based on Wavelet tree Quantization‖, Second IEEE International Symposium on Electronic Commerce and Security, 2009, pp. 218222. [7] Mohammad Ali BaniYounes and AmanJantan, ―Image Encryption Using Block-Based Transformation Algorithm‖, IAENG International Journal of Computer Science, 35:1, IJCS_35_1_03, 2008. [8] Naida.H.Nazmudeen, Farsana.F.J., ―A New Method for Satellite Image Security Using DWT-DCT Watermarking and AES Encryption‖, IJIRSET,Volume 3, Special Issue 5, July 2014 [9] Rawan I. Zaghloul, Enas F. Al-Rawashdeh, ―HSV Image Watermarking Scheme Based on Visual Cryptography‖, World Academy of Science, Engineering and Technology Vol:20 2008-0823 [10] Prashant Shinde and Chetan Mohol, ―Copyright Protection for Images on Android Phones‖, IJRET, Volume: 02, Issue: 11, Nov2013 5