A Hybrid Approach for Information Hiding and Encryption using Multiple LSB s Algorithms

A Hybrid Approach for Information Hiding and Encryption using Multiple LSB s Algorithms H.B.Kekre 1, Tanuja Sarode 2 and Pallavi Halarnkar 3 1 Senior Professor, MPSTME, NMIMS University, Mumbai 2 Associate Professor, TSEC Mumbai University, Mumbai 3 PhD Research Scholar, MPSTME, NMIMS University, Mumbai Abstract A number of techniques are available in literature to provide security to digital images, these techniques give their best to Information Hiding, Image Scrambling and Image Encryption. In this paper we have proposed a Hybrid Approach to secure digital images. The proposed framework is a combination of Information Hiding and Image Encryption. For Information Hiding, four different methods of Multiple LSB s Algorithm are used and evaluated. A number of parameters are also used to evaluate the proposed framework. Experimental results show a good performance. Keywords: Information Hiding, Image Steganography, Image Scrambling, Image Encryption 1. INTRODUCTION Image security is a very important issue nowadays. Image security includes different aspects like Information Hiding, Image Encryption etc. All these are different ways of providing security to digital images. Kaliappan proposed an Information Hiding Technique for JPEG Compressed images[1]. The color image is firstly converted into 1D signal in red, green and blue components, audibly masked frequencies in the 1D signal are determined for each segment. Information hiding is carried out by modifying the spectral power at a pair of commonly occurring masked frequencies. To enable the transmission of stego image, it is compressed via the JPEG coding technique. Experimental results gives a good quality stego image, the method is also very simple to implement. Data retrieval at low level compression and low level noise levels is possible. Higher payload is possible at the cost of quality of the stego image. The drawback of LSB insertion method is that it is easily detected with statistical analysis like RS and chi-square methods. To overcome the drawback Hong Juan et al. proposed a steganography algorithm against the statistical analysis [2]. In this method two sample LSB bits are combined using addition modulo 2 (or m) to form a value which is compared to the part of the secret message. If these two values are equal, then no change is made else add the difference of these two values to the second sample. This process is continued till all the secret message data is embedded. Experimental results show that the proposed method resists the statistical analysis of RS and chi-square test. Ching Yu Yang proposed a color image steganography method based on module substitution [3]. Three types of module substitutions are used to embed secret bits, which is based upon the base value of the blocks under consideration. To improve hiding capacity, the R, G and B component is further encoded by mod u, mod u-v and mod u-v-w substitution. Experimental results show that hiding capacity and PSNR generated are better compared to the existing techniques. Daniela et al. [4] proposed a novel approach of image steganography in YUV color space and its derivative. In this method, RGB cover image is first converted to YUV color space, embedding is carried out only in the V component. The standard LSB method is used for secret message embedding. After embedding in the V component the image is converted back to RGB format. The reason for not using any other component in YUV color space for embedding is that any change in the Y component will linearly propagate directly to the RGB color space. Kekre et al. proposed a multiple LSB s technique in [5], the method makes use of multiple LSB s for hiding the secret message. The method is compared to existing technique of steganography called as PVD. The proposed method was implemented on grayscale images, however it is also extensible over color images. The proposed method gives much more hiding capacity then PVD. Kekre et al. proposed a variation of Multiple LSB s algorithm in [6]. The method named KIMLA and KAMLA, KIMLA makes use of the four MSB s, analyses the binary value, converts it to decimal, increments it with one, then again its binary is reconsidered, the number of one s in it is used as number of LSB s to be replaced in the cover image byte for embedding the secret data. In KAMLA the maximum number of 1 s are considered for replacing the cover image byte for embedding the secret data. Experimental results show a very huge increase in the hiding capacity of the cover image. Volume 3, Issue 6, June 2014 Page 42

Another variation of Multiple LSB s Algorithm (KMLA) was proposed by Kekre et al. in [7]. In this method, while embedding the secret data and checking for cover image byte, if the byte is between 240 to 255, four LSB s can be used for embedding the secret data, the fifth LSB is checked with the message bit to be embedded if it matches with the message bit then a fifth bit is utilized for embedding the secret data. If the value of the cover image byte is between 224 and 239, then again the message bit is checked with the fourth LSB, if it matches then it is utilized for embedding the secret data, else only 3 bits are used for embedding the secret data. Rest of the steps are the same as KMLA. Experimental results show a good increase in the hiding capacity of the cover image. The drawback of the traditional Arnold transform is that it applies only to a square image, this limitation is overcome by Min Li et al. in [8]. The paper proposed an image scrambling method which can be applied to a non square image by dividing the image into multiple square areas and then scrambles each region. Experimental results show better security of the proposed method which provides a safe and a reliable transmission. Kekre et al. proposed an Image scrambling method based on the concept of Relative Prime [9]. Two numbers are said to be relative Prime to each other if they do not have any factors in common except one. Using the concept of relative prime, and finding the correlation between the first row and successive prime rows and placing that row next to first one, that has minimum correlation. The same process is repeated for columns. In this way the image is scrambled. The positions with respect to rows and columns is used as a key for descrambling the same. R-Prime Shuffle technique was traditionally applied over the entire image, it was further extended to blocks in the image of different sizes in [10]. Every block had a relative prime with respect to row and column. This together of all the blocks made the key which can be kept secret for descrambling. Compared to the earlier version of R-Prime, this particular method provides more security. Yicong Zhou et al. proposed a new scrambling technique based on a sequence produced using shift registers [11]. The keys used for scrambling, r the number of shift operations and the distance parameter p, with the help of which it is possible to generate different m-sequences allows user with a wide variety of options of m-sequence to be used for image scrambling. This provides a more secure method and difficult to decode the scrambled image. The proposed algorithm can be applied to 2-D or 3-D images in one step. The method is also robust against attacks such as filters and noise attacks. Mohammed et al. proposed an Image encryption technique called as XLLS [12], it consists of two main parts: encryption/decryption algorithm and ciphered key. In diffusion stage XOR operation is used to modify the pixel values, for encryption purpose Lagrange Process (LP) and Least Square Process (LSP) is used. Decryption of the image is just the reverse operation. Two different options for keys are possible, the proposed system uses a key whose length is 192 bits which is expanded using key expansion algorithm AES -192. The second approach makes use of an image as a key and makes use of CBI key expansion algorithm to expand. Sanfu Wang et al. proposed a new Image Scrambling technique based on folding transform [13]. The folding transform is an orthogonal matrix which allows the image to be folded either up-down or left right. The repeated folding of the image using this procedure scrambles the image. Experimental results show that the method has a good hiding ability with lesser computation burden. The method is also robust under common attacks. Image scrambling techniques are applied over digital images so that the image content becomes meaningless. Dimitri et al. proposed an extension of one dimensional scrambling technique based on discrete Prolate Spheroidal Sequences (DPSS) to two dimensional [14]. The DPSS is optimal in energy concentration in a given frequency sub band. The idea of inverse transformation in finite integer domain is applied to image scrambling by Zhang Ruihong et al. in [15]. In involves both gray and position transformation. The paper also analyzes the security and the scrambling cycle. Experimental results show that the method has a good scrambling effect, a long scrambling cycle. Random numbers are generated using Discrete distributions, they are scaled in the range of 0 to 255 so as to make them suitable for a digital Images. Using the concept of MOD operator and its inverse, image is encrypted. Experimental results show a good quality encryption, histogram analysis is also plotted so as to show the flat histogram obtained from the encrypted image. This method is proposed by Kekre et al. in [16]. Volume 3, Issue 6, June 2014 Page 43

Shuhong Li et al. proposed a fast scrambling algorithm using multi-dimensional orthogonal transform in [17]. The sequence produced using orthogonal transform can be used for image scrambling and descrambling. The proposed method has a high security, as the keys used for scrambling is the one used for generating the sequence. The drawback of the method is that it has a high time complexity, to overcome the said drawback, serial and parallel fast algorithms are proposed. An image is also used as a key image. Experimental results show that this concept can greatly reduce the time complexity. A novel method based on information hiding of two-dimensional images in fractional Fourier domains is proposed in [18]. Firstly image is random shifted using the Jigsaw transformation, then a pixel scrambling algorithm based on Arnold transformation is applied. Then this scrambled image is iteratively encrypted in the fractional Fourier domains using the fractional orders which are chosen randomly. The proposed method has a huge key space thus enhancing the security level. 2. PROPOSED HYBRID FRAMEWORK FOR IMAGE SECURITY In this paper, a Hybrid framework for Information Hiding and Image Encryption is proposed. Following are the steps of the Embedding process 1) Take a Image of size x bytes 2) Apply R-Prime Shuffle Technique to scramble it, A scrambled image is obtained 3) Using KMLA or MKMLA or KIMLA or KAMLA Multiple LSB s Information Hiding technique hide the required secret message, in this paper, an ATM card image is used, a scrambled stego image is obtained. 4) Descramble the scrambled stego image by applying R-Prime shuffle, an innocent stego image is obtained. Figure. 1. Embedding Stage The retrieving process is as follows 1) Take the innocent stego image 2) Apply the R-Prime Shuffle technique, a scrambled stego image is obtained 3) Apply the retrieval algorithm and obtain the secret message image hidden in it. Figure. 2. Retrieving Stage Advantage of the proposed framework is that, if some intruder tries to retrieve the secret message image from the innocent stego image which is being transferred he will get an encrypted message image, which will be difficult to Volume 3, Issue 6, June 2014 Page 44

decode. The results for the same are shown below 2.1 Kekre s Multiple LSB s Algorithm (KMLA) Principle: By analyzing the intensity of every pixel the no of bits that could be embedded be decided so as to make the change imperceptible to the human eye. This is done by embedding more number of bits in high intensity value pixels and less number of bits for low intensity pixels. [5] Every pixel value in this image is analyzed and the following checking process is employed 1. If the value of the pixel say g i, is in the range 240 g i 255 then we embed 4 bits of secret data into the 4 LSB s of the pixel. This can be done by observing the first 4 Most Significant Bits (MSB s). If they are all 1 s then the remaining 4 LSB s can be used for embedding data. 2. If the value of g i (First 3 MSB s are all 1 s), is in the range 224 g i 239 then we embed 3 bits of secret data into the 3 LSB s of the pixel. 3. If the value of g i (First 2 MSB s are all 1 s), is in the range 192 g i 223 then we embed 2 bits of secret data into the 2 LSB s of the pixel. 4. And in all other cases for the values in the range 0 g i 191 we embed 1 bit of secret data in to 1 LSB of the pixel. Similarly, we can retrieve the secret data from the gray values of the stego image by again checking the first four MSB s of the pixel value and retrieve the embedded data. 2.2 Modified Kekre s Multiple LSB s Algorithm (MKMLA) The original Kekre s Multiple LSB s method made use of upto four LSB s for embedding information. The number of LSB s selected depends on the pixel value to control the error. In this Modified Kekre s Mulitple LSB s Algorithm (MKMLA) we have modified this to include fifth LSB matching to increase capacity further [7]. Every pixel value in this image is analyzed and the following checking process is employed 1. If the value of the pixel say g i, is in the range 240 g i 255, then we check for the message bit to be embedded, if it is 1 then we utilize the fifth bit of the pixel value. If the message bit is not 1 then we embed 4 bits of secret data into the 4 LSB s of the pixel. This can be done by observing the first 4 Most Significant Bits (MSB s). If they are all 1 s then the remaining 4 LSB s can be used for embedding data. 2. If the value of g i (First 3 MSB s are all 1 s), is in the range 224 g i 239 then we check whether the message bit to be embedded is 0 then we utilize 4 bits of the pixel value. If the message bit is not 0 then embed 3 bits of secret data into the 3 LSB s of the pixel. 3. If the value of g i (First 2 MSB s are all 1 s), is in the range 192 g i 223 then we embed 2 bits of secret data into the 2 LSB s of the pixel. 4. And in all other cases for the values in the range 0 g i 191 we embed 1 bit of secret data in to 1 LSB of the pixel. 5. The embedding process maintains a matrix to keep a track of the pixels where 5 bits are utilized for embedding process. This helps in the retrieving of the secret message. The retrieving process is very simple by observing the most significant bits and using the matrix maintained. 2.3 Kekre s Improved Multiple LSB Algorithm (KIMLA) To Increase the hiding Capacity, KIMLA algorithm was proposed by Kekre et al. [6]. This Algorithm is based upon the four MSB s in the image byte. The four MSB s of the image byte is first converted to decimal, then a one is added to this decimal, this new decimal is then converted to 4 bit binary, the number of one s in this binary equivalent is used to decide the number of LSB bits to be used in the image byte for hiding the secret message. One exceptional case is the number 15 which is not changed and used as it is. 2.4 Kekre s Advanced Multiple LSB Algorithm (KAMLA) An improvement over KIMLA is KAMLA, in the above algorithm the number of one s in the new decimal decides the number of LSB s to be used for embedding purpose. However, in some cases it may happen that the original byte decimal equivalent s binary may have extra one s then the new decimal, so in this algorithm a max of both is considered for hiding the secret message.[6] 3. EXPERIMENTAL RESULTS For Experimental purpose five 24 bit color images of size 256x256 were used, an ATM card image was used as a secret message for embedding. Volume 3, Issue 6, June 2014 Page 45

Volume 3, Issue 6, June 2014 ISSN 2319-4847 3.1 Kekre s Multiple LSB s Algorithm(KMLA) (a) Original Image (b) Scrambled Image (c) Scrambled Image Stego (d) Stego Image Figure 3. (a) Message Image (b) Encrypted Message Image (c) Retrieved Message Image Figure 4. Figure 3(a) shows the original image, (b) shows the scrambled image obtained by R-Prime Shuffle technique, (c) shows the scrambled stego image after hiding the secret message in it and (d) shows the innocent stego image obtained by descrambling the scrambled stego image using KMLA method Figure 4(a) shows the message image used as a secret message for hiding, (b) shows the encrypted image obtained if some intruder tries to retrieve the secret message from the innocent stego image. And (c) shows the retrieved message image from the scrambled stego image. 3.2 Modified Kekre s Multiple LSB s Algorithm(MKMLA) (a) Original Image (b) Scrambled Image (c) Scrambled Image Stego (d) Stego Image Figure 5. (a) Message Image (b) Encrypted Message Image (c) Retrieved Message Image Figure 6. Figure 5(a) shows the original image, (b) shows the scrambled image obtained by R-Prime Shuffle technique, (c) shows the scrambled stego image after hiding the secret message in it and (d) shows the innocent stego image obtained by descrambling the scrambled stego image MKMLA method Figure 6(a) shows the message image used as a secret message for hiding, (b) shows the encrypted image obtained if some intruder tries to retrieve the secret message from the innocent stego image. And (c) shows the retrieved message image from the scrambled stego image. Volume 3, Issue 6, June 2014 Page 46

Volume 3, Issue 6, June 2014 ISSN 2319-4847 3.3 Kekre s Improved Multiple LSB s Algorithm(KIMLA) (a) Original Image (b) Scrambled Image (c) Scrambled Stego Image (d) Stego Image Figure 7. (a) Message Image (b) Encrypted Message Image (c) Retrieved Message Image Figure 8. Figure 7(a) shows the original image, (b) shows the scrambled image obtained by R-Prime Shuffle technique, (c) shows the scrambled stego image after hiding the secret message in it and (d) shows the innocent stego image obtained by descrambling the scrambled stego image KIMLA method Figure 8(a) shows the message image used as a secret message for hiding, (b) shows the encrypted image obtained if some intruder tries to retrieve the secret message from the innocent stego image. And (c) shows the retrieved message image from the scrambled stego image. 3.4 Kekre s Advanced Multiple LSB s Algorithm(KAMLA) (a) Original Image (b) Scrambled Image (c) Scrambled Image Stego (d) Stego Image Figure 9. (a) Message Image (b) Encrypted Message Image (c) Retrieved Message Image Figure 10. Figure 9(a) shows the original image, (b) shows the scrambled image obtained by R-Prime Shuffle technique, (c) shows the scrambled stego image after hiding the secret message in it and (d) shows the innocent stego image obtained by descrambling the scrambled stego image KAMLA method Figure 10(a) shows the message image used as a secret message for hiding, (b) shows the encrypted image obtained if some intruder tries to retrieve the secret message from the innocent stego image. And (c) shows the retrieved message image from the scrambled stego image. Volume 3, Issue 6, June 2014 Page 47

Table 1:Values of Average correlation between rows and columns of original image and scrambled image, Average Moving Distance and Distance Scrambling factor between original image and scrambled image Lena Image Rows: 0.8454 Cols: 0.7005 KMLA MKMLA KIMLA KAMLA Avg Corr Rows 0.3069 Avg Corr Cols 0.2687 AMD(Max) 360.62 AMD 136.48 DSF 0.378 Kutub Image Rows: 0.2784 Cols: 0.3896 KMLA MKMLA KIMLA KAMLA Avg Corr Rows 0.1816 Avg Corr Cols 0.1865 AMD(Max) 360.62 AMD 133.58 DSF 0.370 Vegetable Image Rows: 0.5176 Cols: 0.4994 KMLA MKMLA KIMLA KAMLA Avg Corr Rows 0.1989 Avg Corr Cols 0.1901 AMD(Max) 360.6245 AMD 132.1023 DSF 0.3663 Baboon Image Rows: 0.4355 Cols: 0.5276 KMLA MKMLA KIMLA KAMLA Avg Corr Rows 0.1854 Avg Corr Cols 0.1916 AMD(Max) 360.62 AMD 130.44 DSF 0.3617 Fruits Image Rows: 0.6203 Cols: 0.6317 KMLA MKMLA KIMLA KAMLA Avg Corr Rows 0.1975 Avg Corr Cols 0.2139 AMD(Max) 360.62 AMD 133.26 DSF 0.3696 Table 2: Values of RMSE, PSNR, between original Image and Stego image Lena KMLA MKMLA KIMLA KAMLA RMSE 1.2724 2.0006 1.9070 1.9885 PSNR 46.038 42.107 42.523 42.160 Kutub RMSE 1.4270 2.2549 1.9138 1.9943 PSNR 45.042 41.068 42.492 42.135 Vegetable RMSE 2.0819 3.0569 2.1490 2.1680 PSNR 41.761 38.425 41.486 41.409 Baboon RMSE 1.0453 1.8014 1.7457 1.8632 PSNR 47.745 43.018 43.291 42.725 Fruits RMSE 1.6000 2.4120 1.9412 1.9874 PSNR 44.048 40.483 42.369 42.165 Volume 3, Issue 6, June 2014 Page 48

Table 3: Values of PAFCPV and NPCR, between original Message Image and Scrambled Message image Lena KMLA MKMLA KIMLA KAMLA PAFCPV 0.4156 0.4157 0.4163 0.4161 NPCR 98.834 99.179 98.359 98.270 Kutub PAFCPV 0.4145 0.4080 0.4172 0.4181 NPCR 98.613 99.088 98.125 98.359 Vegetable PAFCPV 0.4168 0.4143 0.4185 0.4118 NPCR 98.868 99.227 98.540 98.397 Baboon PAFCPV 0.4182 0.4148 0.4163 0.4162 NPCR 98.942 99.157 98.605 98.315 Fruits PAFCPV 0.4156 0.4164 0.4174 0.4130 NPCR 98.829 99.326 98.324 98.354 Figure 11. Embedding Capacity in bits for KMLA, MKMLA, KIMLA and KAMLA 4. CONCLUSION For Experimental analysis, a number of parameters are used to evaluate the proposed framework. The framework makes use of four different variations of Multiple LSB s Algorithm for Information Hiding and R-Prime shuffle technique for Image scrambling. In a scrambled image, it is necessary that the correlation between the rows and columns be minimum or reduced as that found in the original image. Average row and column correlation is calculated in the original image and the scrambled image, from the results obtained it is clear that the correlation between the rows and columns are reduced, minimum obtained in fruits image with a row reduction of 68.17 % and column reduction of 66.14%. To evaluate how good is the scrambling technique, parameters called as Average Moving Distance (AMD), AMD-Max and Distance Scrambling factor is calculated, the maximum obtained in Lena image. To analyze the quality of the stego image, RMSE and PSNR is found out between the original image and innocent stego image. The minimum RMSE is obtained in baboon image in KMLA. To evaluate the quality of the encrypted message image obtained by the framework two parameters Peak Average fractional Change in Pixel Value (PAFCPV) [16] and Number of Pixel Change Rate (NPCR) [12] are used. The highest PAFCPV is obtained in baboon image for KMLA method. As there were different Multiple LSB s techniques used, Embedding capacity was also measured the highest capacity in bits been obtained in Lena for KAMLA method which can be seen from Figure 11. References [1] Gopalan, Kaliappan, An image steganography implementation for JPEG-compressed images. In International Symposium on IEEE Communications and Information Technologies ISCIT'07, pp. 739-744, 2007. [2] Zhang, Hong-Juan, and Hong-Jun Tang. A Novel image steganography algorithm against statistical analysis In Proceedings of IEEE Machine Learning and Cybernetics, vol. 7, pp. 3884-3888. 2007. Volume 3, Issue 6, June 2014 Page 49

[3] Yang, Ching-Yu Color image steganography based on module substitutions, In Proceedings of IEEE Intelligent Information Hiding and Multimedia Signal Processing, IIHMSP 2007. vol. 2, pp. 118-121. 2007. [4] Stanescu, Daniela, Mircea Stratulat, Voicu Groza, Ioana Ghergulescu, and Daniel Borca Steganography in YUV color space. In International Workshop on IEEE Robotic and Sensors Environments, ROSE 2007, pp. 1-4. 2007. [5] Kekre, H. B., Archana Athawale, and Pallavi N. Halarnkar Increased Capacity of Information Hiding in LSB's Method for Text and Image. International Journal of Electrical, Computer & Systems Engineering 2(4), pp. 246-249, 2008. [6] Kekre, H. B., Archana A. Athawale, and Uttara A. Athawale Increased cover capacity using advanced multiple LSB algorithms. In Proceedings of the International Conference & Workshop on Emerging Trends in Technology, pp. 25-31. ACM, 2011. [7] Kekre, H. B., Archana Athawale, and Pallavi N. Halarnkar. Performance evaluation of pixel value differencing and Kekre's modified algorithm for information hiding in images. In Proceedings of the International Conference on Advances in Computing, Communication and Control, pp. 342-346. ACM, 2009. [8] Li, Min, Ting Liang, and Yu-jie He. Arnold Transform Based Image Scrambling Method. In Proceedings of Multimedia Tecnhnology (ICMT 2013). pp.1309-1316. 2013. [9] Kekre, H. B., Tanuja Sarode, and Pallavi Halarnkar. Image Scrambling using R-Prime Shuffle. International Journal of Advanced Research in Electrical, Electronics and Instrumentation Engineering, IJAREEIE, 2(8), pp. 4070-4076. 2013 [10] Kekre, H. B., Tanuja Sarode, and Pallavi Halarnkar. Image Scrambling Using R-Prime Shuffle on Image and Image Blocks. International Journal of Advanced Research in Computer and Communication Engineering 3(2), pp. 5471-5476, 2014. [11] Zhou, Yicong, K. A. R. E. N. Panetta, and Sos Agaian. An image scrambling algorithm using parameter bases M- sequences. In Proceedings of IEEE Machine Learning and Cybernetics, 7, pp.3695-3698. 2008. [12] Shreef, Mohammed A., and Haider K. Hoomod. "Image Encryption Using Lagrange-Least Squares Interpolation." International Journal of Advanced Computer Science and Information Technology (IJACSIT) 2, pp. 35-55. 2013. [13] Wang, Sanfu, Yuying Zheng, and Zhongshe Gao. A new image scrambling method through folding transform. In Proceedings of IEEE Computer Application and System Modeling (ICCASM), 2, pp. V2-395. 2010. [14] Van De Ville, Dimitri, Wilfried Philips, Rik Van de Walle, and Ignace Lemahieu. Image scrambling without bandwidth expansion. IEEE Transactions on Circuits and Systems for Video Technology, 14(6) pp. 892-897.2004. [15] Ruihong, Zhang, and Yu Zhichao. Image Scrambling Encryption Algorithm Based on Limited Integer Domain. In International Symposium on, Intelligence Information Processing and Trusted Computing (IPTC), pp. 693-696. 2010. [16] Kekre, H. B., Tanuja Sarode, and Pallavi Halarnkar. Performance Evaluation of Digital Image Encryption using Discrete Random Distributions and MOD operator. IOSR Journal of Computer Engineering(IOSR-JCE) 16(2), (V) pp. 54-68, 2014. [17] Li, Shuhong, Junmin Wang, and Xing Gao. The fast realization of image scrambling algorithm using multidimensional orthogonal transform. In Proceedings of IEEE Image and Signal Processing, CISP'08. (3), pp. 47-51, 2008. [18] Liu, Shi, and John T. Sheridan. Optical Information Hiding by Combining Image Scrambling Techniques in Fractional Fourier Domains. In Proceedings of Irish Signal and Systems Conference, pp. 249-254. 2001. AUTHOR s Dr. H. B. Kekre has received B.E (Hons.) in Telecomm Engineering from Jabalpur University in 1958, M.Tech (Industrial Electronics) from IIT Bombay in 1960, M.S.Engg. (Electrical Engg.) from University of Ottawa, Canada in 1965 and Ph.D. (System Identification) from IIT Bombayin 1970. He has worked as Faculty of Electrical Engg. and then HOD Computer Science and Engg. at IIT Bombay. After serving IIT for 35 years he retired in 1995. After retirement from IIT, for 13 years he was working as a professor and head in the Department of Computer Engg. and Vice Principal at Thadomal Shahani Engineering. College, Mumbai. Now he is Senior Professor at MPSTME, SVKM s NMIMS University. He has guided 17 Ph.Ds, more than 100 M.E./M.Tech and several B.E./ B.Tech projects, while in IIT and TSEC. His areas of interest are Digital Signal processing, Image Processing and Computer Networking. He has more than 450 papers in National / International Journals and Conferences to his credit. He was Senior Member of IEEE. Presently He is Fellow of IETE, Life Member of ISTE and Senior Member of International Association of Computer Science and Information Technology (IACSIT). Recently fifteen students working under his guidance have received best paper awards. Currently eight research scholars working under his guidance have been awarded Ph. D. by NMIMS (Deemed to be University). At present eight research scholars are pursuing Ph.D. program under his guidance. Dr. Tanuja K. Sarode has received M.E. (Computer Engineering) degree from Mumbai University in 2004, Ph.D. from Mukesh Patel School of Technology, Management and Engg. SVKM s NMIMS University, Vile-Parle (W), Mumbai, INDIA. She has more than 11 years of experience in teaching. Currently working as Assistant Professor in Dept. of Computer Engineering at Thadomal Shahani Engineering College, Mumbai. She is member of International Association of Engineers (IAENG) and International Association of Computer Science and Information Technology (IACSIT). Her areas of interest Volume 3, Issue 6, June 2014 Page 50

are Image Processing, Signal Processing and Computer Graphics. She has 150 papers in National /International Conferences/journal to her credit. Ms. Pallavi N.Halarnkar has received M.E. (Computer Engineering) degree from Mumbai University in 2010, currently persuing her Ph.D. from Mukesh Patel School of Technology, Management and Engg. SVKM s NMIMS University, Vile-Parle (W), Mumbai, INDIA. She has more than 8 years of experience in teaching. Currently working as Assistant Professor in Dept. of Computer Engineering at Mukesh Patel School of Technology, Management and Engg. SVKM s NMIMS University, Vile-Parle (W), Mumbai. She has 27 papers in National /International Conferences/journal to her credit Volume 3, Issue 6, June 2014 Page 51