A robust watermarking scheme against contrasts and scaling ...€¦ · A robust watermarking scheme against contrasts and scaling for PocktNeuro Project Chokri CHEMAK (1, 2), Nabil

A robust watermarking scheme against contrasts and scaling for PocktNeuro Project

Chokri CHEMAK (1, 2), Nabil ELMARZOUQI (2), Mohamed Salim BOUHLEL (1) and Jean Christophe LAPAYRE (2)

(1) Research Unit: Sciences and Technologies of Image and Telecommunications Higher Institute of Biotechnology of Sfax-TUNISIA

(2) Computer Sciences Laboratory of Franche-Comte (L. I. F. C) Franche-Comte University - FRANCE

[email protected], [cchemak, nabil.elmarzouqi, jean-christophe.lapayre]@lifc.univ-fcomte.fr

Abstract This paper presents a robust watermarking scheme of medical information security and terminal mobile phone adaptation for PocketNeuro project. The later term refers to a project created for the service of neurological diseases. It consists of transmitting information about patients « Desk of Patients » to a doctor’s mobile phone when he is visiting or examining his patient. Our scheme provides image robustness for both image Gray-level transformation and scaling to adapt them to the doctor’s mobile phone. Experiments carried out on a database of 20- 256× 256 pixel-sized medical images show that our watermarking scheme is robust to both scaling and Gray-level transformations. For the purpose of increasing the image watermarking robustness against attacks of an image transmission and to perform a large number of bits to hide into images, we encode with turbo code image-embedded information. Results demonstrate that fidelity can also be improved by this image watermarking scheme. Keywords: Feature Points, Delaunay Tesselation, RGB colour image, Turbo code, Robustness, Fidelity, collaboration, awareness. 1. Introduction With the support of European/INTERREGIII funding (with Switzerland), the Distributed System team in the Computer Laboratory of Franche-Comte (LIFC) established a new project with Besançon Institute of Sciences and Technologies of Information (ISTI) coordinated by the University Hospital Center (CHU) of Lausanne entitled PocketNeuro. This Project is responsible for mobility between terminals (in Cooperative Tele-neurology). It allows practitioners to use telecommunication technologies to provide medical deontology and facilitate distant diagnosis [1]. This project is created for the services of neurological diseases [2]. The aim of this project is to transmit diagnostic neuronal images containing information about patients to a doctor’s mobile phone when he is visiting his patient (see figure1).

Figure1. Example of Mobile Phone transmission for PocketNeuro Project PocketNeuro project needs to provide security of medical information and adaptation of neuronal images to the doctor’s mobile phone. We propose to use watermarking [3] as solution to preserve the security of medical information for the PocketNeuro project. Image watermarking allows doctors or personnel to hide invisible and robust medical information about patient inside images with a secret key. To adapt transmitted images to the doctor’s mobile phone, we must scale and transform these images from colour to Gray-level. On the one hand, to resist image watermarked against scaling, the embedding bits information is done by extracting Feature Points of the image and performing a Delaunay Tesselation on the set of points [4]. On the other hand, information is embedded in blue channel of RGB colour image to guaranteed good fidelity of image after image transformation from colour to gray-level.

ICGST-GVIP Journal, ISSN: 1687-398X , Volume 8, Issue 1, June 2008

33

mailto:[email protected]

For the purpose of increasing the image watermarking robustness against attacks of an image transmission and to ensure large data payloads, we encode with turbo code image-embedded bits information [5]. In our paper, images are extracted from DICOM library and we use Matlab and its library as software to experiment our approaches and to validate our results. We use as technical requirements (Hardware) a Modern computing terminals and a cooperative platform made in neurology services. The paper is organized as follows: In section 2, we first give an overview of previous countermeasures proposed in the literature to resist image watermarked against scaling. Furthermore, we explain the original contribution of the approach based on the choice of the Feature Points extraction and the Delaunay Tesselation. We follow that by explaining the metrics evaluation used in our results simulations. In Section 3, we describe the main steps relevant to our watermark embedding and detecting scheme. In Section 4, we attempt to account for the idea of watermarking colour image. We will present the original contribution of our scheme and we will describe the main steps of our embedding scheme. In section 5, we list some simulation results showing that even after scaling and transforming colour image to Gray-level; all 1000 bits are correctly detected in at least 90% of watermarked images while preserving image fidelity after watermarking scheme. In section 6, we present NeuroPocket project and the use of watermarking process in a virtual collaborative session. The detection of embedded marks is assured by detection process when treated medical image arrived to collaborators. Finally, we comprise some concluding remarks concerning the advantages of our watermarking scheme. 2. Use of the Delaunay Tesselation and evaluation metrics 2. 1.Use of the Delaunay Tesselation In this section, we present the outline of classical methods to resist image watermarked against scaling beginning from the methods resorting to the original image until the methods using Feature Points of the image. Dong et al. [6] as well as Davoine et al. [7] propose a method resorting to the original image to compensate for the geometrical deformations such as scaling. It consists of applying regular triangular Tesselation on both original and watermarked images. The compensation for geometrical distortions is performed by slightly shifting different vertices of the attacked image Tesselation to minimize the quadratic error between each original triangle and the corresponding attacked triangle. But, this method is very efficient only in the case of minor scaling. Another solution consists in embedding the watermark in a geometrical invariant subspace. Coltuc and Bolon [8] suggest using histogram specification to hide a watermark invariant

to geometrical distortions. Ruanaidh and Pun [9], and Lin et al. [10] propose a watermarking scheme based on the Fourier–Mellin transform [11. In practice, this solution can be implemented for simple affine, but the problems of approximation due to the discrete nature of the images, plus the reduction of the embedding space make the watermark weakly resistant to low-pass filtering and lossy compression. The last class of resynchronization techniques (also called second generation schemes [12]) uses image content to recover the watermark after geometrical transformations. The method suggested by Bas et al. [13] consists in extracting Feature Points of the image using a Harris detector [14] and performing a Delaunay Tesselation on the set of points. We select feature point detectors to represent an image content descriptor. It is well known that Feature Points detectors find salient points in natural images. These points are often located near corners and edges of the image.The set of feature point permits image partitioning using a Delaunay Tesselation. Some images from our database, on which we are carried out experiments and doing the extraction of Feature Points and the Tesselation, are shown in figure 2. We focus our choice on Delaunay Tesselation because it has local properties: if a vertex disappears, the Tesselation is only modified on connected triangles; each vertex is associated with a stability area in which the Tesselation is not modified [15] when the vertex is moving inside this area. Consequently, we don’t lose bits information embedded inside each triangle of the Tesselation after watermarked images scaling.

Figure2. Some figures from the database of 20 medical images

2. 2. Evaluation Metrics 2.2.1. The correlation detector We use correlation to declare if the watermark is present or absent. The original bits information embedded into image belongs to a bank of 500 bits medical formations. Each bits information of this bank is supposed to be the bits information embedded in the image. We can determine the inserted information by computing the correlation of the extracted bits information with every element of the bank [16]. Indeed, the inserted bits medical information should have the highest correlation with the extracted one. The equation of the correlation is follow:

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛∑∑ −∑∑ −

∑∑ −−=

m,n²Β)m, n(Β

m,n²Α)m, n(Α

mnΒ)m, n(ΒΑ)m, n(Α

)B,A(Cor (1)


34

Where, A and B are matrices or vectors of the same size and are the means of the matrices A and B.

Correlation between the information extracted and

the embedded one

Information extracted, decoded and Wiener

prediction

Feature point detector

Delaunay Tesselation

2.2.2. The Peak Signal to Noise Ratio (PSNR) The Peak Signal to Noise Ratio (PSNR) present the distortion measured in Decibels (db) caused by the watermarking in image. It is derived from the Mean Square Error (MSE). The later term refers to a metric used to quantify the distortion in images generated by the digital watermark after the embedding process [17] [18]. The equations are follows:

( )nm

IIMSE

n

1i

m

1j

2*ijij∑ ∑ −

= = =

Where, I and I* are respectively the original image and the watermarked one.

)MSE/255(log10)MSE/X(log10PSNR

210

2max10

=

=

Where, X max is the maximum of luminescence in the image. 3. Our robust watermarking scheme against image scaling We present in this section our method for image watermarking robust against scaling 3.1. The embedding Scheme Firstly, bits information embedding uses references provided by feature point detectors using Harris detector [15]. Next, a Delaunay Tesselation is employed to decompose the image into a set of disjoint triangles. The information embedding is performed by embedding bits information inside each triangle of the Tesselation. Pixels in each vertex are substituted with the turbo code-coded bit information in the Second Low Symbol Byte (LSB2). After that we scale watermarked image with report 0.8 to adapt it to the terminal portable phone receiver (See figure 3). 3.2. The detection Scheme The detection of Feature Points using our improved Harris detector is applied to the scaled image. We obtain a set of points. After that, a Delaunay Tesselation is made. The detection of information is performed in each triangle. Then, we use Wiener detection with its statistics to eliminate components provided by the original image. Finally, a correlation between original bits information and the extracted and the decoded one is made.

The principle of the detection process is detailed in figure 4. A, B Original image

4. Our technique robust against colour to indexed Gray-scale images transformation To adapt image to the terminal, we need to transform colour image to an indexed Gray-level one. We start the embedding process by treating each colour channel separately, as in [19], [20] and [21]. An alternative approach to colour image watermarking has been advanced by Kutter, Jordan, and Bossen [22], who proposed embedding the watermark by modifying a selected set of pixel values in the blue channel, because the human eye is less sensitive to changes in this band.

Figure 4. Principle of the detection process

Watermarked image

Scaling

Delaunay Tesselation

Feature point detector

1

2

3

4

5

(2)

(3)

3

4

5

6

Watermarked image

TK

TI

Information coded with Turbo code

Original Mark

LSB2 substitution of convex pixels with bits

information

Add Psychvisual weighting

1

2

Figure 3. Principle of the embedding process


35

Our suggested watermarking method is shown below in figure 5.

Figure 5. (a) Colour image devised in three channels (Blue, Green, and Red), (b) Image watermarked by LSB2 substitution of selected pixel values in the blue channel.), (c) Image transformed from coloured to an indexed Gray-level

5. Simulation results 5.1. Robustness against scaling and Gray-level image transformation This section displays first, the experimental results carried out on a database of 500 bits information that contain the original medical information in which the extracted and decoded bits information are tested. The test technique is made by correlation between the extracted, decoded bits information and the dictionary. The following figures give a summary of the test simulations results obtained after scaling and image Gray-level transformation. From the one side; figure 6 demonstrates that we have maxima of correlation between the original bits information and the extracted one. As such result validates the robustness of our watermarking scheme and proves the perfect detection of information from the medical image after image scaling. In this way, we explore the advantages of Feature Point and the Delaunay Tesselation to keep the watermark invariant in each vertex. From another side; the bits information owner is perfectly detected from indexed figure after Gray-level image transformation (see figure 7). This result proves the advantages of our watermarking scheme to keep robustness of images rather after transformation from colour to Gray-level. Furthermore, the incorporation of turbo code in the formatting of the watermark increases the number of bits to hide in order to achieve higher data payloads because the number of repetitions of each bit of the watermark decreases at the same proportion [23], [24]. From our simulation results, all the 1000 bits are correctly detected in at least 90% from the 20 medical image banks.

5.2. Fidelity of watermarked images 5.2.1. Fidelity of watermarked images after scaling R (c) It is quite relevant to evaluate the fidelity of our images after watermarking. In fact, in PocketNeuro project, we need to clearly give content of diagnostic image for all doctors’ portable phones. From the figure 8, we can observe values of PSNRs for 20, 256× 256 sized-pixel medical images after watermarking and scaling process. These values are always higher than 30dB. This makes it obvious that the image quality is good and these watermarked images technique is powerful to keep image fidelity after scaling.

0 100 200 300 400 500-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Cor

rela

tion

Figure 6. Succeeded detection of information after

correlation between extracted and decoded bits information and the dictionary after image scaling

0 50 100 150 200 250 300 350 400 450 500-0.2

0

0.2

0.4

0.6

0.8

Cor

rela

tion

Figure 7. Succeeded detection of information after correlation

between extracted, decoded bits information and the dictionary after image Gray-level transformation

0

10

20

30

40

50

60

70

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29

30 Medical images

PSN

R

Series1

njb

Figure 8. Mean values of PSNRs for 20 test watermarked images after scaling

G

B

(a)

(b)

3

2

5

4

Embedding mark after a turbo code-coding in

the blue channel

Mark extracted and decoded

Correlation between extracted and decoded

mark and the dictionary

1

Gray-level image transformation


36

5.2.2. Fidelity of images against watermark sizes after scaling From the figure 9, we can observe the mean values of PSNRs for 20, 256× 256 sized-pixel medical images after scaling process with different watermark sizes: 1500 Bits, 1000 Bits, 800 Bits, 600 Bits and 500 Bits. These mean of values are above 30dB. Indeed, we have a good fidelity of images after image watermarking with different length of watermark even after scaling of images.

0

10

20

30

40

5060

PSNR

1500Bits

1000Bits

800Bits

600Bits

500Bits

Sizes of watermark

Figure 9. Mean values of PSNRs for 20 test watermarked images

with different watermark sizes after scaling

5.2.3. Fidelity of watermarked images after transformation from colour to Gray-level We often need to apply a Gray-level image transformation to reduce the information in image for transmission between terminals. We tried to test the fidelity of different indexed image after transformations by measuring the PSNRs for the different images. We worked at three medical images from our images database. These images are shown in figure 10 and results of different values of PSNR are shown in Table 1. We can remark that the PSNRs are higher than 50 dB. So, fidelity is improved after different transformations from colour image to Gray-level indexed images. 5.2.4. Fidelity against watermark sizes after image transformation from colour to Gray-level indexed image In figure 11, we can show the mean values of PSNRs for 20 medical images after image transformation from colour to Gray-level indexed image with different watermark sizes: 1500 Bits, 1000 Bits, 800 Bits, 600 Bits and 500 Bits. We remark the good fidelity of images with different watermark sizes. 5.3. Fidelity of our proposed technique against noises It is quite relevant to evaluate the robustness of the suggested method against noises. We have tested our new approach using 5 different noises generations and by modifying variances at each time. From the figure 12, we can observe values of PSNRs that are always higher than 30 dB. This makes it obvious that the image quality is good and this new watermarked image

scheme is powerful to keep image fidelity even after noises attack.

Table 1: Mean values of PSNRs of four test watermarked images

after Gray-level image transformation

Medical images PSNRs Image 1 65,73 Image 2 56, 91 Image 3 67, 53

Medical image 1 Medical image 2 Medical image 3

Figure 10. Some figures from the database of colour neuronal images which they are used in our simulation to validate fidelity of our watermarking technique after Gray-level images transformation

0

20

40

60

80

PSN

R

1500 Bits 1000 Bits 800 Bits 600 Bits 500 BitsWatermark Sizes

Figure11. Mean values of PSNRs for images of figure 10 with different watermark sizes after transformation from colour to Gray-

level indexed images

28

28,5

29

29,5

30

30,5

31

PSNR

Gaussiannoise0,01

Salt andpaper

noise 0,015

Specklenoise 0,

02

Figure 12. Mean values of PSNRs for 20 test images watermarked

and attacked by different types of noises


37

6. Watermarking in collaborative session In collaborative environment, several actions are produced by the various collaborators leading to a sharp inclusion of awareness in a multi-user environment. Collaborative Virtual Environment (CVE) is characterized by a shared space permitting to have all participants in the same place in order to give illusion to be in the same place. Every actor has possibility of interacting not only with other participants present in the multi-user environment, but also with the software components of virtual environment. ProcketNeuro is a collaborative environment for neurologists dedicated to treatment of medical images in virtual examination room. The PocketNeuro platform allow to several doctors to treat a patient’s images in real-time session by sharing image documents and discussing the neurological pathology presented for patient. As presented in figure 13, we have applied the watermarking process to the medical image during collaborative real time session where doctor, in an examination virtual room, sent a watermarked image to other doctors in order to inspect medical image. All doctors are aware about embedded information added by the sender and then the identification of mark’s owner is achieved by applying a detection process. Doctors present in real-time session can interact with each other to give their standpoint. Medical image with embedded information allow also keeping the integrity of diagnosis realized in collaborative session. Events established during diagnosis are preserved as well as the treated areas and decisions taken during treatment. Conclusion The aim of this paper is to outline adaptability of terminals to the watermarking image transmission for PocketNeuro Project in the field of emergency neurology. PocketNeuro will allow the neurologist to access relevant medical information for any given patient in the "patient file”. It should allow patient-related medical information to be consulted and modified in a secure mode. Adaptation of watermarking images to the terminals is made by scaling image and transform colour image to an indexed Gray-level. We demonstrate by some simulation results that our proposed techniques for watermarked images scheme for terminal adaptability purposes are robust against scale and Gray-level transformations. Our watermarking approaches use powerful error-correcting turbo codes to improve resistance of watermarking after image transmission and to perform a large number of bits embedded inside medical images. Simulation result shows that after scale and Gray-level image transformation, all bits are detected in at least 90% from operated images. Furthermore, for different tested attacks the perceptual metric incorporated the PSNR allow us to keep perfect values of images distortions above 30 dB. Consequently, we

have a good perceptual fidelity in images after watermarking process. Tests performed on this first PocketNeuro release have given good results, and allow us to hope for a large experimentation between French and Switzerland hospitals using secured connections between the Tele-neurology platforms in Lausanne (Switzerland) and terminal mobiles phone of doctors in French.

Figure 13. Multi-diffusion of a watermarked image in a

collaborative session 7. Acknowledgements The authors extend their thanks both to the Besançon Institute of Sciences and Technologies of Information (ISTI), for initiating this project and the European Union for financing this project as part of the INTERREGIII program in collaboration with Swiss partners (Vaud University Hospital Center of Lausanne, and EPFL Lausanne). References [1] X. KONG, R. FENG,” Watermarking medical

signals for Telemedicine”. IEEE Transactions on Information Technology in Biomedicine, 2001, Vol.5, Issue 3, p. 195-201.

[2] E. GARCIA, H.GUYENNET, J-C LAPAYRE and T. MOULIN,” Adaptive Tele-Application for Remote Neurology Diagnosis”. Journal of Telemedicine and e-Health, 2005, vol. 11, no. 6, p. 692-702.

[3] J. Bernarding, A. Thiel, A. Grzesik, “A JAVA-based DICOM server with integration of clinical findings and DICOM-conform data encryption”. International Journal of Medical Informatics 2001; 64:429-438.

[4] J-L. Dugelay, S. Roche, C. Rey, and G. Doërr, “Still-Image Watermarking Robust to Local Geometric Distortions”, IEEE Transactions on Image Processing, September 2006, Vol. 15, No. 9, pp. 2831-2842.

[5] C. Chemak, M-S Bouhlel and J-C Lapayre, "Algorithme de Tatouage Robuste et Aveugle pour la Déontologie et le Transfert des


38

Informations Médicales: Le Tatouage Combinée ATRADTIM", Brevet d’Invention déposés à l’INNORPI, Decembre 2006, SN06448.

[6] P. Dong, J. Brankov, N. Galatsanos, and Y. Yang, “Geometric robust watermarking through mesh model based correction,” in Proc. IEEE Int. Conf. Image Processing, Sep. 2002, vol. 3, pp. 493–496.

[7] F. Davoine, P. Bas, P.-A. Hebert, and J.-M. Chassery, “Watermarking et résistance aux déformations géométriques,” in Proc. Compression et Représentation des Signaux Audiovisuels, Jun. 1999.

[8] D. Coltuc and P. Bolon, “Robust watermarking by histogram specification,” in Proc. IEEE Int. Conf. Image Processing, Oct. 1999, vol. 2, pp. 236–239.

[9] J. J. K. Ó Ruanaidh and T. Pun, “Rotation, scale and translation invariant digital image watermarking,” Signal Process., vol. 66, no. 3, pp. 303–317, May 1998.

[10] C. Y. Lin, M. Wu, J. A. Bloom, I. J. Cox, M. L. Miller, and Y. M. Lui, “Rotation, scale, and translation-resilient public watermarking for images,” Proc. SPIE—Security and Watermarking of Multimedia Contents II, vol. 3971, pp. 90–98, Jan. 2000.

[11] F. Lin and R. D. Brandt, “Towards absolute invariants of images under translation, rotation and dilatation,” Pattern Recognit. Lett., vol. 14, no. 5, pp. 369–379, May 1993.

[12] M. Kutter, S. K. Bhattacharjee, and T. Ebrahimi, “Towards second generation watermarking schemes,” in Proc. IEEE Int. Conf. Image Processing, Oct. 1999, vol. 1, no. 10, pp. 320–323.

[13] P. Bas, J.-M. Chassery, and B. Macq, “Geometrically invariant watermarking using Feature Points,” IEEE Trans. Image Process., vol. 11, no. 9, pp. 1014–1028, Sep. 2002.

[14] C. Harris and M. Stephen, “A combined corner and edge detector,” in Proc. 4th Alvey Vision Conf., 1988, vol. 15, pp. 147–151.

[15] E. Bertin, S. Marchand-Maillet and J.-M. Chassery, “Optimization in Voronoi Diagrams”, Norwell, MA: Kluwer, 1994, pp. 209–216.

[16] C. Chemak, J-C Lapayre, M-S. Bouhlel,” New Watermarking Scheme for Security and Transmission of Medical Images for PocketNeuro Project”, Radioengineering.journal, Special Issue: Advanced Digital Signal Processing 2007, December 2007, Vol.16, pp. 58-63.

[17] T.Aura, “Partical invisibility in digital communication”. In proceeding of the Workshop on information Hiding, number 1174 in Lecture Notes in Computer Science, Springer Verlag, Cambridge, England, Mai 1996.

[18] S.Craver, N.Memon, B-L.Yeo, “Resolving Rightful Ownerships with Invisible Watermarking Techniques: Limitations, Attacks and Implic-ations”, IEEE Journal on selected areas in communications, VoL 16, N0. 4, Mai 1998.

[19] M. T. Sandford, II, T. G. Handel and J. N. Bradley, “The data embedding method”, in Proc. SPIE Photonics East, vol. 2615–2628, Oct. 23–26, 1995.

[20] N. Nikolaidis and I. Pitas, “Robust image watermarking in the spatial domain”, Signal Processing, vol. 66, no. 3, pp. 385–403, May 1998.

[21] M. Barni, Member, IEEE, Franco Bartolini, Member, IEEE, and Alessandro Piva, “Multichannel Watermarking of Color Images”, IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, VOL. 12, NO. 3, MARCH 2002.

[22] M. Kutter, F. Jordan, and F. Bossen, “Digital signature of color images using amplitude modulation in Storage and Retrieval for Image and Video Databases V”, SPIE Vol. 3022, San Jose, CA, Feb. 1997, pp. 518–526.

[23] C. Berrou and A. Glavieux, "Near optimum error correcting coding and decoding: Turbo codes", IEEE Trans. Comm., pp. 1261 - 1271, Oct. 1996.

[24] J.R.Hernández, J.-F. Delaigle and B. Macq, “Improving Data Hiding by Using Convolutional Codes and Soft-Decision Decoding”, Proceedings of SPIE - Security and Watermarking of Multimedia Contents II, pp. 24-48, San Jose, USA, January 2000.


39

Biographies Pr. Mohamed Salim BOUHLEL was born in Sfax, Tunisia, in 1955. He received the Ph.D. degree from the University of Lyon, French, in 1981 and the H.D.R. from the University of Sfax, Tunisia, in 2003. Currently, he is a professor and Head of the Biomedical imagery

Department in the Higher Institute of Biotechnology of Sfax, Tunisia. In 1999, he received the golden medal with the special mention from the jury in the First International Meeting of Invention, Innovation and Technology (Dubai). He is currently the Vice President of the Tunisian Association of the Experts in Imagery and President of the Tunisian Association of the Experts in Information Technology and Telecommunication. He is the Editor in Chief of the international Journal of Electronic, Technology of Information and Telecommunication (JETIT), Chairman of the international conference: Sciences of Electronic, Technologies of Information and Telecommunication: (SETIT 2003, SETIT 2004, SETIT 2005 and SETIT 2007). His research interests include image processing in the field of Telemedicine applications.

Chokri CHEMAK was born in CHEBBA, Tunisia, in 1978. He received the Master degree from the National Engineering School of Sfax, Tunisia, in June 2005. Since September 2005, he was been an Assistant Professor of Computer Science in the High Institute of

Computer Sciences, Mahdia, Tunisia. Currently, he is studding for the Ph.D. degree. He is a member of the Distributed System Team in the Computer Science Laboratory of Franche-Comte (L.I.F.C.), French. He is also, a member of the R.U.: Sciences and Technologies of Image and Telecommunications (SETIT), Tunisia. He is a member of the organisation committee of the IEEE International Conference: Sciences of Electronic, Technologies of Information and Telecommunication: (SETIT 2005 and SETIT 2007). His research interests include image processing (including Watermarking and data hiding and information Security).

Nabil ELMARZOUQI was born in MEKNES, Morocco, in 1981. He received the Master degree from Val de Marne University of Paris, Creteil, in July 2005. Since September 2005, he was recruited by French Ministry of National Education, Research and Technology (MENRT), ministerial scholarship to

prepare his Ph.D degree. Currently he is a member of Multimedia Collaborative Work Group in the Distributed Systems and Networks team (SDR) of Computer Science Laboratory of Franche-Comté University (LIFC), France. He is moreover part-time lecturer in Franche-Comté University and biomedical engineering school of Besançon, ISIFC (Institute of Engineers of Franche-Comté). He is also a member of the organization committee of the IEEE International Conference of E-MEDIcal SYStems (E-MEDISYS 2007). His research interests include Distributed Systems and Collaborative work applied to Telemedicine environments and Telediagnosis applications (including medical image processing in collaborative platforms).

Pr. Jean-Christophe LAPAYRE was born in Paris (Antony 92), in 1963. He was an Assistant Professor from 1998 until august 2003. Currently, he is working in the Distributed Systems & Network Research Team, in the Multimedia Cooperative Work Group. He was been a Professor since

2003 and the Head of the Computer Science Department of Franche-Comte University, since January 2003. He is the Head of the Distributed Algorithmic for Tele-Applications Research Group in the Computer Science Laboratory of Franche-Comte (LIFC), since 2002. He was the Chairman of the IEEE International Conferences: E-MEDIcal SYStem 2007 (E-MEDISYS 2007), DFMA(Distributed Framework for Multimedia Applications) 2005, 2006 and 2007. His research interests include Distributed Systems in the field of Telediagnosis and Telemedicine applications. He was the leader of the European projects Teneci (Teleneurolgie collaborative) and Decopreme (Detection collaborative precoce du melanoma). J.-C. Lapayre is a member of the SET (Société Européenne de Télémédecine). Since January 2008, He is the Director of the Computer Scie

nce Lab. IFC.

L


40

Documents

A robust watermarking scheme against contrasts and scaling ...€¦ · A robust watermarking scheme against contrasts and scaling for PocktNeuro Project Chokri CHEMAK (1, 2), Nabil