Professor in Signal and Image Processing at the Faculty of Science of the University of Sfax Tunisia. Member of the Laboratory of Electronics and Information Technologies at the National School of Engineering of sfax
The International Arab Journal of Information Technology, 2023
Although recent advances in Deep Learning (DL) algorithms have been developed in many Computer Vi... more Although recent advances in Deep Learning (DL) algorithms have been developed in many Computer Vision (CV) tasks with a high accuracy level, detecting humans in video streams is still a challenging problem. Several studies have, therefore, focused on the regularisation techniques to prevent the overfitting problem which is one of the most fundamental issues in the Machine Learning (ML) area. Likewise, this paper thoroughly examines these techniques, suggesting an improved You Only Look Once (YOLO)v3-tiny based on a modified neural network and an adjusted hyperparameters file configuration. The obtained experimental results, which are validated on two experimental tests, show that the proposed method is more effective than the YOLOv3-tiny predecessor model . The first test which includes only the data augmentation techniques indicates that the proposed approach reaches higher accuracy rates than the original YOLOv3-tiny model. Indeed, Visual Object Classes (VOC) test dataset accuracy...
2019 International Conference on Internet of Things, Embedded Systems and Communications (IINTEC), 2019
Several military files and medical research reports do not tolerate any type of distortion and ar... more Several military files and medical research reports do not tolerate any type of distortion and are at the same time secret to owners. In order to preserve source information and guarantee data security, lossless video compression with encryption seems to be the best alternative. In this work, cryptography was grafted with lossless compression to ensure a high image quality and secure the transmission of such data. Thus, the insertion of the encryption module in the lossless video compression chain appears to be better than the application of compression and encryption separately in terms of computing time. Our developed method reached about 30% gain of total bits and around 75% improvement of the compression ratio proving promising results.
2014 1st International Conference on Advanced Technologies for Signal and Image Processing (ATSIP), 2014
ABSTRACT Currently, research is focused on selective encryption to protect movies against attacks... more ABSTRACT Currently, research is focused on selective encryption to protect movies against attacks transmitted over a public network. This choice is taken to ensure the security of compressed data without adding an important supplemental computational time for the encryption. So the encryption module is inserted into the video compression chain. In this paper we present an encryption method based chaos inserted in the H.264/AVC chain specially for the motion vectors signs. The results will be deducted according to the values of PSNR (Peak Signal to Noise Ratio), SSIM (Structural SIMilarity) and the encryption rate. Motion vectors signs encryption (MVSE) seems being interesting to lose the visual quality of the video for Inter predicted frames since movies are sensible to the motion sign.
ABSTRACT It is possible to determine the absolute intensity of a light signal recorded by a video... more ABSTRACT It is possible to determine the absolute intensity of a light signal recorded by a video camera with automatic gain by analyzing the background part of the recorded images. The method is based on the determination of the histogram of the gray levels of the dark pixels whose inverse variance appears to be proportional to the camera gain in a large domain of light intensity. An example is given by images recorded by a charge coupled device camera in a space experiment where the light transmittivity is used as a local probe for temperature measurements.
World Academy of Science, Engineering and Technology, International Journal of Computer, Electrical, Automation, Control and Information Engineering, 2004
he compression is an important area of research in image processing and has been widely studied i... more he compression is an important area of research in image processing and has been widely studied in the last decades. Compression algorithms can restitute the exact information (Lossless methods) or introduce a little distortion (lossy methods). The Lossless image compression is designed to reduce or remove the image’s redundancy. They have weak compression ratio but benefit from an exact reconstruction of the image. The purpose of lossy image compression is to minimize the number of bits needed to represent an image without introducing an important degradation. Natural images have an important redundancy among the space-image pixels. A linear transformation is applied to minimize redundancy in the images since it can decorrelate pixel values in the transform domain. The international standard for still image compression, called Joint Photographic Experts Group or JPEG standard [1]-[2], uses the Discrete Cosine Transform (DCT) [3]-[4]. The JPEG2000 [5] uses The Discrete Wavelet Trans...
Nowday, UAV is new technologie among the next generation monitoring system which has the followin... more Nowday, UAV is new technologie among the next generation monitoring system which has the following properties: efficient, real time, secure and reliable. UAV occupied a great importance in many fileds especially national defense in order to protect the human. In fact, the size of transferred data is a basic demand for realtime video communication, partically for military service since the bandwidth limitation in UAV networks.Exprimental results show that the H264 standard outperforms MPEG 4 (Motion Picture Expert Group) encoder in termes of PSNR (Peak signal to noise ratio). The H264’PSNR gain is 6.88% better than MPEG 4. In addition, the H264 standard did not exceed 20 Mbit\s for wireless video transmission.
International Conference on Internet of Things, Embedded Systems and Communications, IEEE IINTEC2019, 2019
Several military files and medical research reports do not tolerate any type of distortion and ar... more Several military files and medical research reports do not tolerate any type of distortion and are at the same time secret to owners. In order to preserve source information and guarantee data security, lossless video compression with encryption seems to be the best alternative. In this work, cryptography was grafted with lossless compression to ensure a high image quality and secure the transmission of such data. Thus, the insertion of the encryption module in the lossless video compression chain appears to be better than the application of compression and encryption separately in terms of computing time. Our developed method reached about 30% gain of total bits and around 75% improvement of the compression ratio proving promising results.
Lossless video coding improves the coding efficiency of the standard H.264/AVC. In this work, we ... more Lossless video coding improves the coding efficiency of the standard H.264/AVC. In this work, we adjusted the standard H.264/AVC to make it conservative. We removed the quantification process of the standard H.264/AVC coding chain because this block is the first responsible for the degradation of the reconstructed video. In the framework of the lossless compression, we compared two entropy codings: CABAC (context-based adaptive binary arithmetic coding) and CAVLC (context-based adaptive variable length coding). The simulation results showed that CABAC reduces the total bit of the coded sequence, but increased the execution time compared to CAVLC. Thanks to the lossless video coding, the gain performance reached from 15 to 19 %.
The 17th international conference on Sciences and Technics of Automatic control and computer engineering, IEEE STA’2016, 2016
The utilization of human visual system characteristics in measuring the quality of images and vid... more The utilization of human visual system characteristics in measuring the quality of images and video sequences becomes a very important axis of research. The best objective criterion is as close as possible to the judgment of the human being. Our idea is to establish a pretreatment to existing measure with complete reference (PSNR "Peak Signal-to-Noise Ratio") for the H.264/MPEG-4 (Motion Picture Expert Group) AVC (Advanced Video Coding) standard. We perform a pretreatment using SURF (Speeded Up Robust Features) saliency map and a CSF filter with distance of 500 mm to improve the performance of these metrics. For the evaluation of our proposed model, the database Video subjective "Live" is used to measure the correlation. Performance indicators, namely Pearson (PLCC), the Spearman coefficient (SROCC) and mean squared prediction error (RMSE) indicate that our idea gives better results for the distortion caused by H264 video encoding.
Lossy video coding can be achieved by using the H.264 Advanced Video Coding. Some methods have be... more Lossy video coding can be achieved by using the H.264 Advanced Video Coding. Some methods have been proposed to transform this coder into a lossless one. The hierarchical lossless video coding structure with new intra algorithm prediction was proposed to reach this goal. In this paper we proposed some enhancements of this method to improve the coding efficiency. The simulation results show that the proposed enhancements reduce the total bit of the coded sequence and the execution time.
Nowday, UAV is new technologie among the next generation monitoring system which has the followin... more Nowday, UAV is new technologie among the next generation monitoring system which has the following properties: efficient, real time, secure and reliable. UAV occupied a great importance in many fileds especially national defense in order to protect the human. In fact, the size of transferred data is a basic demand for realtime video communication, partically for military service since the bandwidth limitation in UAV networks.Exprimental results show that the H264 standard outperforms MPEG 4 (Motion Picture Expert Group) encoder in termes of PSNR (Peak signal to noise ratio). The H264’PSNR gain is 6.88% better than MPEG 4. In addition, the H264 standard did not exceed 20 Mbit\\s for wireless video transmission.
Peak Signal to Noise Ratio (PSNR) and Structural SIMilarity (SSIM) are metrics initially used to ... more Peak Signal to Noise Ratio (PSNR) and Structural SIMilarity (SSIM) are metrics initially used to evaluate the visual quality of compressed images or sequences compared to the original ones. By analogy to compressed sequences, researchers use these metrics to evaluate the degradation of encrypted sequences. Video encryption algorithms target a maximum scrambling so that their contents become imperceptible to the human visual system. The distortion of PSNR and SSIM values comes from both compression and encryption. The use of these metrics to measure the degradation of joint compressed and encrypted sequences cannot give us a precise evaluation of the distortion. For a better evaluation, a Contrast Sensitivity Function (CSF) metric was used. This article aims to provide a perceptual evaluation of the encryption effect for the H.264 Advanced Video Coding compressed and encrypted sequences, using the CSF metric. The visual quality of the encrypted video is degraded and proven bad from different viewing distances.
The 360-deg or virtual reality video can capture all sides around the observer at the same time a... more The 360-deg or virtual reality video can capture all sides around the observer at the same time and provides the freedom to select any part of the surroundings for display using a head mounted display in the case of a single user or smart TV, curved screen, and even with video projectors placed in a 360-deg projection room in the case of multiusers. In order to encode these spherical videos with standard codecs, a projection step is necessary to transform the original 3-D space scene into a regular 2-D video sequence. Thus different geometric shapes can be used as equirectangular projection (ERP), cubemap projection (CMP), segmented sphere projection (SSP), etc. The proposed model presents the 360-deg video on eight square faces, six faces for the equatorial part, and two faces for the top and bottom views. This shape can ensure a better projection of the video thanks to the size ratio of the lateral faces, which is very close to the size ratio of televisions and video projectors used in 360-deg rooms. Quality metrics comparison shows average gains when compared with ERP and SSP models and the encoding time is hugely reduced when compared with CMP model.
The International Arab Journal of Information Technology, 2023
Although recent advances in Deep Learning (DL) algorithms have been developed in many Computer Vi... more Although recent advances in Deep Learning (DL) algorithms have been developed in many Computer Vision (CV) tasks with a high accuracy level, detecting humans in video streams is still a challenging problem. Several studies have, therefore, focused on the regularisation techniques to prevent the overfitting problem which is one of the most fundamental issues in the Machine Learning (ML) area. Likewise, this paper thoroughly examines these techniques, suggesting an improved You Only Look Once (YOLO)v3-tiny based on a modified neural network and an adjusted hyperparameters file configuration. The obtained experimental results, which are validated on two experimental tests, show that the proposed method is more effective than the YOLOv3-tiny predecessor model . The first test which includes only the data augmentation techniques indicates that the proposed approach reaches higher accuracy rates than the original YOLOv3-tiny model. Indeed, Visual Object Classes (VOC) test dataset accuracy...
2019 International Conference on Internet of Things, Embedded Systems and Communications (IINTEC), 2019
Several military files and medical research reports do not tolerate any type of distortion and ar... more Several military files and medical research reports do not tolerate any type of distortion and are at the same time secret to owners. In order to preserve source information and guarantee data security, lossless video compression with encryption seems to be the best alternative. In this work, cryptography was grafted with lossless compression to ensure a high image quality and secure the transmission of such data. Thus, the insertion of the encryption module in the lossless video compression chain appears to be better than the application of compression and encryption separately in terms of computing time. Our developed method reached about 30% gain of total bits and around 75% improvement of the compression ratio proving promising results.
2014 1st International Conference on Advanced Technologies for Signal and Image Processing (ATSIP), 2014
ABSTRACT Currently, research is focused on selective encryption to protect movies against attacks... more ABSTRACT Currently, research is focused on selective encryption to protect movies against attacks transmitted over a public network. This choice is taken to ensure the security of compressed data without adding an important supplemental computational time for the encryption. So the encryption module is inserted into the video compression chain. In this paper we present an encryption method based chaos inserted in the H.264/AVC chain specially for the motion vectors signs. The results will be deducted according to the values of PSNR (Peak Signal to Noise Ratio), SSIM (Structural SIMilarity) and the encryption rate. Motion vectors signs encryption (MVSE) seems being interesting to lose the visual quality of the video for Inter predicted frames since movies are sensible to the motion sign.
ABSTRACT It is possible to determine the absolute intensity of a light signal recorded by a video... more ABSTRACT It is possible to determine the absolute intensity of a light signal recorded by a video camera with automatic gain by analyzing the background part of the recorded images. The method is based on the determination of the histogram of the gray levels of the dark pixels whose inverse variance appears to be proportional to the camera gain in a large domain of light intensity. An example is given by images recorded by a charge coupled device camera in a space experiment where the light transmittivity is used as a local probe for temperature measurements.
World Academy of Science, Engineering and Technology, International Journal of Computer, Electrical, Automation, Control and Information Engineering, 2004
he compression is an important area of research in image processing and has been widely studied i... more he compression is an important area of research in image processing and has been widely studied in the last decades. Compression algorithms can restitute the exact information (Lossless methods) or introduce a little distortion (lossy methods). The Lossless image compression is designed to reduce or remove the image’s redundancy. They have weak compression ratio but benefit from an exact reconstruction of the image. The purpose of lossy image compression is to minimize the number of bits needed to represent an image without introducing an important degradation. Natural images have an important redundancy among the space-image pixels. A linear transformation is applied to minimize redundancy in the images since it can decorrelate pixel values in the transform domain. The international standard for still image compression, called Joint Photographic Experts Group or JPEG standard [1]-[2], uses the Discrete Cosine Transform (DCT) [3]-[4]. The JPEG2000 [5] uses The Discrete Wavelet Trans...
Nowday, UAV is new technologie among the next generation monitoring system which has the followin... more Nowday, UAV is new technologie among the next generation monitoring system which has the following properties: efficient, real time, secure and reliable. UAV occupied a great importance in many fileds especially national defense in order to protect the human. In fact, the size of transferred data is a basic demand for realtime video communication, partically for military service since the bandwidth limitation in UAV networks.Exprimental results show that the H264 standard outperforms MPEG 4 (Motion Picture Expert Group) encoder in termes of PSNR (Peak signal to noise ratio). The H264’PSNR gain is 6.88% better than MPEG 4. In addition, the H264 standard did not exceed 20 Mbit\s for wireless video transmission.
International Conference on Internet of Things, Embedded Systems and Communications, IEEE IINTEC2019, 2019
Several military files and medical research reports do not tolerate any type of distortion and ar... more Several military files and medical research reports do not tolerate any type of distortion and are at the same time secret to owners. In order to preserve source information and guarantee data security, lossless video compression with encryption seems to be the best alternative. In this work, cryptography was grafted with lossless compression to ensure a high image quality and secure the transmission of such data. Thus, the insertion of the encryption module in the lossless video compression chain appears to be better than the application of compression and encryption separately in terms of computing time. Our developed method reached about 30% gain of total bits and around 75% improvement of the compression ratio proving promising results.
Lossless video coding improves the coding efficiency of the standard H.264/AVC. In this work, we ... more Lossless video coding improves the coding efficiency of the standard H.264/AVC. In this work, we adjusted the standard H.264/AVC to make it conservative. We removed the quantification process of the standard H.264/AVC coding chain because this block is the first responsible for the degradation of the reconstructed video. In the framework of the lossless compression, we compared two entropy codings: CABAC (context-based adaptive binary arithmetic coding) and CAVLC (context-based adaptive variable length coding). The simulation results showed that CABAC reduces the total bit of the coded sequence, but increased the execution time compared to CAVLC. Thanks to the lossless video coding, the gain performance reached from 15 to 19 %.
The 17th international conference on Sciences and Technics of Automatic control and computer engineering, IEEE STA’2016, 2016
The utilization of human visual system characteristics in measuring the quality of images and vid... more The utilization of human visual system characteristics in measuring the quality of images and video sequences becomes a very important axis of research. The best objective criterion is as close as possible to the judgment of the human being. Our idea is to establish a pretreatment to existing measure with complete reference (PSNR "Peak Signal-to-Noise Ratio") for the H.264/MPEG-4 (Motion Picture Expert Group) AVC (Advanced Video Coding) standard. We perform a pretreatment using SURF (Speeded Up Robust Features) saliency map and a CSF filter with distance of 500 mm to improve the performance of these metrics. For the evaluation of our proposed model, the database Video subjective "Live" is used to measure the correlation. Performance indicators, namely Pearson (PLCC), the Spearman coefficient (SROCC) and mean squared prediction error (RMSE) indicate that our idea gives better results for the distortion caused by H264 video encoding.
Lossy video coding can be achieved by using the H.264 Advanced Video Coding. Some methods have be... more Lossy video coding can be achieved by using the H.264 Advanced Video Coding. Some methods have been proposed to transform this coder into a lossless one. The hierarchical lossless video coding structure with new intra algorithm prediction was proposed to reach this goal. In this paper we proposed some enhancements of this method to improve the coding efficiency. The simulation results show that the proposed enhancements reduce the total bit of the coded sequence and the execution time.
Nowday, UAV is new technologie among the next generation monitoring system which has the followin... more Nowday, UAV is new technologie among the next generation monitoring system which has the following properties: efficient, real time, secure and reliable. UAV occupied a great importance in many fileds especially national defense in order to protect the human. In fact, the size of transferred data is a basic demand for realtime video communication, partically for military service since the bandwidth limitation in UAV networks.Exprimental results show that the H264 standard outperforms MPEG 4 (Motion Picture Expert Group) encoder in termes of PSNR (Peak signal to noise ratio). The H264’PSNR gain is 6.88% better than MPEG 4. In addition, the H264 standard did not exceed 20 Mbit\\s for wireless video transmission.
Peak Signal to Noise Ratio (PSNR) and Structural SIMilarity (SSIM) are metrics initially used to ... more Peak Signal to Noise Ratio (PSNR) and Structural SIMilarity (SSIM) are metrics initially used to evaluate the visual quality of compressed images or sequences compared to the original ones. By analogy to compressed sequences, researchers use these metrics to evaluate the degradation of encrypted sequences. Video encryption algorithms target a maximum scrambling so that their contents become imperceptible to the human visual system. The distortion of PSNR and SSIM values comes from both compression and encryption. The use of these metrics to measure the degradation of joint compressed and encrypted sequences cannot give us a precise evaluation of the distortion. For a better evaluation, a Contrast Sensitivity Function (CSF) metric was used. This article aims to provide a perceptual evaluation of the encryption effect for the H.264 Advanced Video Coding compressed and encrypted sequences, using the CSF metric. The visual quality of the encrypted video is degraded and proven bad from different viewing distances.
The 360-deg or virtual reality video can capture all sides around the observer at the same time a... more The 360-deg or virtual reality video can capture all sides around the observer at the same time and provides the freedom to select any part of the surroundings for display using a head mounted display in the case of a single user or smart TV, curved screen, and even with video projectors placed in a 360-deg projection room in the case of multiusers. In order to encode these spherical videos with standard codecs, a projection step is necessary to transform the original 3-D space scene into a regular 2-D video sequence. Thus different geometric shapes can be used as equirectangular projection (ERP), cubemap projection (CMP), segmented sphere projection (SSP), etc. The proposed model presents the 360-deg video on eight square faces, six faces for the equatorial part, and two faces for the top and bottom views. This shape can ensure a better projection of the video thanks to the size ratio of the lateral faces, which is very close to the size ratio of televisions and video projectors used in 360-deg rooms. Quality metrics comparison shows average gains when compared with ERP and SSP models and the encoding time is hugely reduced when compared with CMP model.
International Image Processing, Applications and Systems Conference IPAS, 2014
In This paper, we propose a new perceptually significant video quality metric to estimate the eff... more In This paper, we propose a new perceptually significant video quality metric to estimate the effect of block coding for standards H.264 AVC and MPEG2. Our method operates in the spatial domain and doesn't require a high complexity of computation. We compare it with Suthaharan's and MSU's techniques by using “LIVE” database. Results indicate that the proposed method outperforms Suthaharan's technique. They also indicate that our method is more effective than MSU's technique for the H.264 AVC and MPEG2 standard with the Spearman Rank Order Correlation Coefficient.
Sciences of Electronic, Technologies of Information and Telecommunications SETIT, Mar 17, 2003
Dans les différentes applications du traitement d’image, les limitations du système visuel humain... more Dans les différentes applications du traitement d’image, les limitations du système visuel humain (SVH) peuvent être exploitées dans le but de l’amélioration des performances en termes de qualité visuelle des images. Même si les exigences spécifiques pour chacune de ces applications sont différentes, l'élément commun est toujours un modèle mathématique de la vision humaine. Sa structure générale est déterminée habituellement par l’étude des effets et des critères psychophysiques du système SVH. De plus l’utilisation du modèle SVH apporte une amélioration importante pour les méthodes subjectives de mesure de la qualité des images. L'objectif principal de la compression de l'image est d’atteindre un haut niveau de compression sans perte de l’information visuelle. Cela implique une distinction entre les informations visibles et invisibles contenues dans une image. Cette distinction est effectuée par l’exploitation d’un modèle visuel humain (SVH) incorporé au sein du processus de compression. L’élément clé d'un tel modèle est "la fonction de sensibilité au contraste" (CSF), basée essentiellement sur la sensibilité réduite du système visuel pour des structures de hautes fréquences spatiales. Son exploitation lors de la phase de quantification, permet la restriction de la troncature des informations contenues dans une image, en d’autres termes la troncature affecte intensivement les informations de hautes fréquences spatiales et assure la préservation des fréquences spatiales visuellement importantes. Par conséquent cette approche permet d’améliorer significativement la qualité visuelle de la compression. Dans ce contexte notre travail consiste à l’exploitation d’un modèle visuel humain à base de la CSF pour l'élaboration d'une nouvelle quantification incorporée au sein de la chaîne de compression basée sur la transformée en ondelettes discrète (TOD) utilisée dans la norme JPEG2000 et permet d’améliorer la qualité visuelle de la compression. Notre nouvelle méthode est évaluée en utilisant le PSNR pondéré communément appelé wPSNR.
Dans notre travail, nous nous sommes intéressés à établir une métrique avec référence complète po... more Dans notre travail, nous nous sommes intéressés à établir une métrique avec référence complète pour la norme H264/AVC pour nous approcher le plus possible du jugement du système visuel humain (SVH). Comme première étape, nous avons travaillé sur la pondération du PSNR classique. En effet, l’être humain n’arrive pas à détecter les dégradations dans les séquences où il y a beaucoup de mouvement. Pour cela, nous avons intégré dans le PSNR l’effet de mouvement des macroblocs 16×16. Chaque macrobloc a un vecteur de mouvement de module V. Nous adoptons en second lieu la métrique NCD (Normalized Color Difference) qui reflète mieux la loi de Weber. Pour cela, il faut passer de l’espace YUV à l’espace Lab. Puis nous calculons la NCD pondérée. Pour avoir une meilleure corrélation avec le SVH, nous avons essayé de filtrer les séquences originales et dégradées avec la Fonction de Sensibilité au Contraste (CSF). Pour cela, nous avons réalisé la chaîne de filtrage avant de calculer la NCD. Pour le filtrage, nous avons essayé trois types de filtres existants. Ces filtres sont réalisés sur la partie luminance seulement. Nous avons réalisé un filtrage de CSF sur les composantes couleurs. Nous avons utilisé un filtre à partir de la littérature.
International Conference on Internet of Things, Embedded Systems and Communications, IEEE IINTEC2019, 2019
Several military files and medical research reports do not tolerate any type of distortion and ar... more Several military files and medical research reports do not tolerate any type of distortion and are at the same time secret to owners. In order to preserve source information and guarantee data security, lossless video compression with encryption seems to be the best alternative. In this work, cryptography was grafted with lossless compression to ensure a high image quality and secure the transmission of such data. Thus, the insertion of the encryption module in the lossless video compression chain appears to be better than the application of compression and encryption separately in terms of computing time. Our developed method reached about 30% gain of total bits and around 75% improvement of the compression ratio proving promising results.
The 17th international conference on Sciences and Technics of Automatic control and computer engineering, IEEE STA’2016, 2016
The utilization of human visual system characteristics in measuring the quality of images and vid... more The utilization of human visual system characteristics in measuring the quality of images and video sequences becomes a very important axis of research. The best objective criterion is as close as possible to the judgment of the human being. Our idea is to establish a pretreatment to existing measure with complete reference (PSNR "Peak Signal-to-Noise Ratio") for the H.264/MPEG-4 (Motion Picture Expert Group) AVC (Advanced Video Coding) standard. We perform a pretreatment using SURF (Speeded Up Robust Features) saliency map and a CSF filter with distance of 500 mm to improve the performance of these metrics. For the evaluation of our proposed model, the database Video subjective "Live" is used to measure the correlation. Performance indicators, namely Pearson (PLCC), the Spearman coefficient (SROCC) and mean squared prediction error (RMSE) indicate that our idea gives better results for the distortion caused by H264 video encoding.
Dans le but de transmettre les données médicales
entre les centres hospitaliers, nous insérons le... more Dans le but de transmettre les données médicales entre les centres hospitaliers, nous insérons les informations médicales de chaque patient dans l’image de son diagnostic. Pour cela, nous appliquons un tatouage sur les images médicales. En effet, le tatouage consiste à insérer une signature (un message) dans l'image et essayer de retrouver cette signature avec le maximum de fidélité possible. Le rôle du tatouage est de respecter la confidentialité des données insérées dans l’image médicale, puisqu’il permet d’insérer les informations introduites dans l'image suivant une clé secrète [11]. Ce papier présente un nouveau schéma de tatouage capable d’augmenter la robustesse du tatouage des images médicales en préservant la qualité de l’image et sa confidentialité. La robustesse des images médicales sera testée contre différentes attaques comme (le bruit additif, le cropping et compression JPEG). Mots clés : Tatouage d’image, turbo code, Corrélation, PSNR, robustesse, SOVA
entre les centres hospitaliers, nous insérons les
informations médicales de chaque patient dans
l... more entre les centres hospitaliers, nous insérons les informations médicales de chaque patient dans l’image de son diagnostic. Pour cela, nous appliquons un tatouage sur les images médicales. En effet, le tatouage consiste à insérer une signature (un message) dans l'image et essayer de retrouver cette signature avec le maximum de fidélité possible. Le rôle du tatouage est de respecter la confidentialité des données insérées dans l’image médicale, puisqu’il permet d’insérer les informations introduites dans l'image suivant une clé secrète [11]. Ce papier présente un nouveau schéma de tatouage capable d’augmenter la robustesse du tatouage des images médicales en préservant la qualité de l’image et sa confidentialité. La robustesse des images médicales sera testée contre différentes attaques comme (le bruit additif, le cropping et compression JPEG). Mots clés : Tatouage d’image, turbo code, Corrélation, PSNR, robustesse, SOVA… 1. Introduction
Uploads
Papers by Fahmi Kammoun
L'objectif principal de la compression de l'image est d’atteindre un haut niveau de compression sans perte de l’information visuelle. Cela implique une distinction entre les informations visibles et invisibles contenues dans une image. Cette distinction est effectuée par l’exploitation d’un modèle visuel humain (SVH) incorporé au sein du processus de compression. L’élément clé d'un tel modèle est "la fonction de sensibilité au contraste" (CSF), basée essentiellement sur la sensibilité réduite du système visuel pour des structures de hautes fréquences spatiales. Son exploitation lors de la phase de quantification, permet la restriction de la troncature des informations contenues dans une image, en d’autres termes la troncature affecte intensivement les informations de hautes fréquences spatiales et assure la préservation des fréquences spatiales visuellement importantes. Par conséquent cette approche permet d’améliorer significativement la qualité visuelle de la compression.
Dans ce contexte notre travail consiste à l’exploitation d’un modèle visuel humain à base de la CSF pour l'élaboration d'une nouvelle quantification incorporée au sein de la chaîne de compression basée sur la transformée en ondelettes discrète (TOD) utilisée dans la norme JPEG2000 et permet d’améliorer la qualité visuelle de la compression. Notre nouvelle méthode est évaluée en utilisant le PSNR pondéré communément appelé wPSNR.
H264/AVC pour nous approcher le plus possible du jugement du système visuel humain (SVH). Comme première étape,
nous avons travaillé sur la pondération du PSNR classique. En effet, l’être humain n’arrive pas à détecter les dégradations
dans les séquences où il y a beaucoup de mouvement. Pour cela, nous avons intégré dans le PSNR l’effet de mouvement des
macroblocs 16×16. Chaque macrobloc a un vecteur de mouvement de module V. Nous adoptons en second lieu la métrique
NCD (Normalized Color Difference) qui reflète mieux la loi de Weber. Pour cela, il faut passer de l’espace YUV à l’espace
Lab. Puis nous calculons la NCD pondérée. Pour avoir une meilleure corrélation avec le SVH, nous avons essayé de filtrer
les séquences originales et dégradées avec la Fonction de Sensibilité au Contraste (CSF). Pour cela, nous avons réalisé la
chaîne de filtrage avant de calculer la NCD. Pour le filtrage, nous avons essayé trois types de filtres existants. Ces filtres sont
réalisés sur la partie luminance seulement. Nous avons réalisé un filtrage de CSF sur les composantes couleurs. Nous avons
utilisé un filtre à partir de la littérature.
entre les centres hospitaliers, nous insérons les
informations médicales de chaque patient dans
l’image de son diagnostic. Pour cela, nous
appliquons un tatouage sur les images médicales.
En effet, le tatouage consiste à insérer une
signature (un message) dans l'image et essayer de
retrouver cette signature avec le maximum de
fidélité possible.
Le rôle du tatouage est de respecter la
confidentialité des données insérées dans l’image
médicale, puisqu’il permet d’insérer les
informations introduites dans l'image suivant une
clé secrète [11].
Ce papier présente un nouveau schéma de
tatouage capable d’augmenter la robustesse du
tatouage des images médicales en préservant la
qualité de l’image et sa confidentialité.
La robustesse des images médicales sera testée
contre différentes attaques comme (le bruit additif,
le cropping et compression JPEG).
Mots clés : Tatouage d’image, turbo code,
Corrélation, PSNR, robustesse, SOVA
informations médicales de chaque patient dans
l’image de son diagnostic. Pour cela, nous
appliquons un tatouage sur les images médicales.
En effet, le tatouage consiste à insérer une
signature (un message) dans l'image et essayer de
retrouver cette signature avec le maximum de
fidélité possible.
Le rôle du tatouage est de respecter la
confidentialité des données insérées dans l’image
médicale, puisqu’il permet d’insérer les
informations introduites dans l'image suivant une
clé secrète [11].
Ce papier présente un nouveau schéma de
tatouage capable d’augmenter la robustesse du
tatouage des images médicales en préservant la
qualité de l’image et sa confidentialité.
La robustesse des images médicales sera testée
contre différentes attaques comme (le bruit additif,
le cropping et compression JPEG).
Mots clés : Tatouage d’image, turbo code,
Corrélation, PSNR, robustesse, SOVA…
1. Introduction