Takwa CHIHAOUI | ESIEE - Academia.edu (original) (raw)
Papers by Takwa CHIHAOUI
Notre travail s’inscrit dans le cadre de la biometrie par la retine. La retine est la couche sens... more Notre travail s’inscrit dans le cadre de la biometrie par la retine. La retine est la couche sensorielle de l’œil, elle presente une texture riche et unique meme chez les jumeaux. Ses proprietes ont fait de la biometrie par la retine un axe de recherche actif. En effet, de nombreuses methodes ont ete proposees pour les differentes etapes de la methode biometrique allant du pretraitement de l’image retinienne a son analyse, en passant par sa caracterisation, afin d’identifier et authentifier un individu. Nous nous interessons dans ces travaux de these, a l’etude, la conception, le developpement et l’evaluation d’une nouvelle methode biometrique basee sur la retine. Notre premiere contribution reside dans la conception d’une methode d’analyse d’image retinienne saine et pathologique, suivie d’une selection d’une region d’interet autour du disque optique. Cette methode ameliore la qualite de l’image retinienne et extrait une region d’interet la plus stable de la retine afin de maintenir une densite d’information satisfaisante, pour assurer une meilleure qualite de reconnaissance. Notre deuxieme contribution porte sur la proposition d’une nouvelle methode d’extraction de caracteristiques locales basee sur le descripteur standard SIFT. Elle applique une nouvelle approche reposant sur la suppression des points d’interet non informatifs extraits par le descripteur standard SIFT. Cette nouvelle methode d’extraction des caracteristiques locales reduit le nombre des points d’interet redondants tout en maintenant la qualite de la description. Nous avons valide, la methode biometrique proposee sur differentes bases comprenant des images saines et pathologiques. Les resultats obtenus montrent des performances encourageantes. Ces resultats indiquent, que la methode que nous avons proposee, localise correctement la region d’interet retinienne. En mode identification, un taux d’identification correcte d’environ 99.8% est atteint. En mode verification, nous avons obtenu un taux FRR de 0.12% quant aux taux FAR et EER (erreur), ils sont de 0%. L’etude comparative a montre que notre methode est plus discriminative que d’autres methodes de l’etat de l’art, notamment celles basees sur la segmentation et l’extraction de l’arbre vasculaire
L’identification des personnes est devenue un service d’exigence croissante pour la maitrise des ... more L’identification des personnes est devenue un service d’exigence croissante pour la maitrise des risques lies au controle d’acces physique a des milieux hautement securises (coffre-fort, base militaire, etc.). De nos jours, les technologies d’authentification biometrique apparaissent comme dominantes pour la reconnaissance automatique des personnes. En effet, des caracteristiques biometriques distinctives et mesurables comme l’iris, l’empreinte digitale, la forme de visage, l’ADN, la retine etc., sont de plus en plus utilisees. La plupart des efforts dans les systemes d'authentification ont tendance a developper des environnements plus surs et difficiles a falsifier. Dans ce contexte, le motif des vaisseaux retinien est unique, universel, invariable dans le temps et tres difficile a falsifier. Il est considere comme un des caracteristiques biometriques les plus efficaces pour l'authentification personnelle. Par ailleurs, en raison des processus d'acquisition modernes, le...
Communication and Signal Processing, Jul 23, 2018
Personal recognition based on retina has been an attractive topic of scientific research. A commo... more Personal recognition based on retina has been an attractive topic of scientific research. A common challenge of retinal identification system is to ensure a high recognition rate while maintaining a low mismatching FMR rate and execution time. In this context, this paper presents a retinal identification system based on a novel local feature description. The proposed system is composed of three stages, firstly we enhance the retinal image and we select a ring around the optical disc as an interest region by using our recently proposed Optical Disc Ring ODR method. Secondly, in order to reduce the mismatching rate and speed up the matching step, we propose in this paper an original alternative local description based on the Remove of Uninformative SIFT Keypoints, that we call SIFT-RUK. Finally, the generalization of Lowe's matching technique (g2NN test) is employed. Experiments on the VARIA database are done to evaluate the performance of our proposed SIFT-RUK feature-based identification system. We show that we obtain a high performance with 99.74% of identification accuracy rate without any mismatching (0% of False Matching Rate FMR) and with a low matching processing time compared to existing identification systems.
2016 13th International Multi-Conference on Systems, Signals & Devices (SSD), 2016
Today, Human recognition, especially based on retina, has been an important and attractive topic ... more Today, Human recognition, especially based on retina, has been an important and attractive topic of scientific research. Most efforts in Biometrics tend to develop more efficient systems which compromise speed and robustness of authentication. In fact, retinal images often suffer from imperfections such as background intensity variation, affine transformations (translation, rotation, scale changes) variations from pattern to other. These defects can seriously affect features extraction in terms of quality and execution time. In this context, in order to overcome these defects, we propose in this paper a novel retinal verification system based on the Speeded Up Robust Features (SURF) extraction. This feature extraction method is so fast and invariant to the affine transformations such as rotation, scale changes and translation. We employ the Optical Disc interest Ring (ODR) method as a preprocessing step in order to further speed up the system and improve the performance. A subset of the VARIA database is used to evaluate the proposed SURF based system. It compromises a high quality with 100% of verification accuracy rate and a time processing very lower than existing verification systems.
2015 International Conference on Image Processing Theory, Tools and Applications (IPTA), 2015
2014 1st International Conference on Advanced Technologies for Signal and Image Processing (ATSIP), 2014
ABSTRACT In recent years, more data are created in digital form allowing easy control over storag... more ABSTRACT In recent years, more data are created in digital form allowing easy control over storage and manipulation due to the technology progress. Unfortunately, this progress may have dragged along a lot of risks, especially the ones related to the security of digital files. In particular, digital forgery becomes a worry for many organizations because it has become easier to create fake images without leaving any obvious perceptual traces of tampering. A specific form of image forgery operation called “copy-move” is considered one of the most difficult problems in the case of forgery detection. For this case, a part of the image is copied and pasted on another location of the same image to conceal undesirable objects in the scene. In this paper, we propose a method that automatically detects duplicated regions in the same image. Duplicated detection is performed by identifying the local characteristics of the images (points of interest) using the Scale Invariant Feature Transform (SIFT) method and by matching between identical features using the Singular Value Decomposition (SVD) method. Obtained results show that our proposed hybrid method is robust to geometrical transformations and is able to detect with high performance duplicated regions.
Notre travail s’inscrit dans le cadre de la biometrie par la retine. La retine est la couche sens... more Notre travail s’inscrit dans le cadre de la biometrie par la retine. La retine est la couche sensorielle de l’œil, elle presente une texture riche et unique meme chez les jumeaux. Ses proprietes ont fait de la biometrie par la retine un axe de recherche actif. En effet, de nombreuses methodes ont ete proposees pour les differentes etapes de la methode biometrique allant du pretraitement de l’image retinienne a son analyse, en passant par sa caracterisation, afin d’identifier et authentifier un individu. Nous nous interessons dans ces travaux de these, a l’etude, la conception, le developpement et l’evaluation d’une nouvelle methode biometrique basee sur la retine. Notre premiere contribution reside dans la conception d’une methode d’analyse d’image retinienne saine et pathologique, suivie d’une selection d’une region d’interet autour du disque optique. Cette methode ameliore la qualite de l’image retinienne et extrait une region d’interet la plus stable de la retine afin de maintenir une densite d’information satisfaisante, pour assurer une meilleure qualite de reconnaissance. Notre deuxieme contribution porte sur la proposition d’une nouvelle methode d’extraction de caracteristiques locales basee sur le descripteur standard SIFT. Elle applique une nouvelle approche reposant sur la suppression des points d’interet non informatifs extraits par le descripteur standard SIFT. Cette nouvelle methode d’extraction des caracteristiques locales reduit le nombre des points d’interet redondants tout en maintenant la qualite de la description. Nous avons valide, la methode biometrique proposee sur differentes bases comprenant des images saines et pathologiques. Les resultats obtenus montrent des performances encourageantes. Ces resultats indiquent, que la methode que nous avons proposee, localise correctement la region d’interet retinienne. En mode identification, un taux d’identification correcte d’environ 99.8% est atteint. En mode verification, nous avons obtenu un taux FRR de 0.12% quant aux taux FAR et EER (erreur), ils sont de 0%. L’etude comparative a montre que notre methode est plus discriminative que d’autres methodes de l’etat de l’art, notamment celles basees sur la segmentation et l’extraction de l’arbre vasculaire
L’identification des personnes est devenue un service d’exigence croissante pour la maitrise des ... more L’identification des personnes est devenue un service d’exigence croissante pour la maitrise des risques lies au controle d’acces physique a des milieux hautement securises (coffre-fort, base militaire, etc.). De nos jours, les technologies d’authentification biometrique apparaissent comme dominantes pour la reconnaissance automatique des personnes. En effet, des caracteristiques biometriques distinctives et mesurables comme l’iris, l’empreinte digitale, la forme de visage, l’ADN, la retine etc., sont de plus en plus utilisees. La plupart des efforts dans les systemes d'authentification ont tendance a developper des environnements plus surs et difficiles a falsifier. Dans ce contexte, le motif des vaisseaux retinien est unique, universel, invariable dans le temps et tres difficile a falsifier. Il est considere comme un des caracteristiques biometriques les plus efficaces pour l'authentification personnelle. Par ailleurs, en raison des processus d'acquisition modernes, le...
Communication and Signal Processing, Jul 23, 2018
Personal recognition based on retina has been an attractive topic of scientific research. A commo... more Personal recognition based on retina has been an attractive topic of scientific research. A common challenge of retinal identification system is to ensure a high recognition rate while maintaining a low mismatching FMR rate and execution time. In this context, this paper presents a retinal identification system based on a novel local feature description. The proposed system is composed of three stages, firstly we enhance the retinal image and we select a ring around the optical disc as an interest region by using our recently proposed Optical Disc Ring ODR method. Secondly, in order to reduce the mismatching rate and speed up the matching step, we propose in this paper an original alternative local description based on the Remove of Uninformative SIFT Keypoints, that we call SIFT-RUK. Finally, the generalization of Lowe's matching technique (g2NN test) is employed. Experiments on the VARIA database are done to evaluate the performance of our proposed SIFT-RUK feature-based identification system. We show that we obtain a high performance with 99.74% of identification accuracy rate without any mismatching (0% of False Matching Rate FMR) and with a low matching processing time compared to existing identification systems.
2016 13th International Multi-Conference on Systems, Signals & Devices (SSD), 2016
Today, Human recognition, especially based on retina, has been an important and attractive topic ... more Today, Human recognition, especially based on retina, has been an important and attractive topic of scientific research. Most efforts in Biometrics tend to develop more efficient systems which compromise speed and robustness of authentication. In fact, retinal images often suffer from imperfections such as background intensity variation, affine transformations (translation, rotation, scale changes) variations from pattern to other. These defects can seriously affect features extraction in terms of quality and execution time. In this context, in order to overcome these defects, we propose in this paper a novel retinal verification system based on the Speeded Up Robust Features (SURF) extraction. This feature extraction method is so fast and invariant to the affine transformations such as rotation, scale changes and translation. We employ the Optical Disc interest Ring (ODR) method as a preprocessing step in order to further speed up the system and improve the performance. A subset of the VARIA database is used to evaluate the proposed SURF based system. It compromises a high quality with 100% of verification accuracy rate and a time processing very lower than existing verification systems.
2015 International Conference on Image Processing Theory, Tools and Applications (IPTA), 2015
2014 1st International Conference on Advanced Technologies for Signal and Image Processing (ATSIP), 2014
ABSTRACT In recent years, more data are created in digital form allowing easy control over storag... more ABSTRACT In recent years, more data are created in digital form allowing easy control over storage and manipulation due to the technology progress. Unfortunately, this progress may have dragged along a lot of risks, especially the ones related to the security of digital files. In particular, digital forgery becomes a worry for many organizations because it has become easier to create fake images without leaving any obvious perceptual traces of tampering. A specific form of image forgery operation called “copy-move” is considered one of the most difficult problems in the case of forgery detection. For this case, a part of the image is copied and pasted on another location of the same image to conceal undesirable objects in the scene. In this paper, we propose a method that automatically detects duplicated regions in the same image. Duplicated detection is performed by identifying the local characteristics of the images (points of interest) using the Scale Invariant Feature Transform (SIFT) method and by matching between identical features using the Singular Value Decomposition (SVD) method. Obtained results show that our proposed hybrid method is robust to geometrical transformations and is able to detect with high performance duplicated regions.