Pub Date : 2008-11-01DOI: 10.1109/IPTA.2008.4743776
Messaoud Bengherabi, L. Mezai, F. Harizi, A. Guessoum, M. Cheriet
Although information fusion in unimodal or multimodal biometric systems can be performed at various levels, integration of the matching score level is the most common approach. Starting from the fact; that the fusion will be efficient if and only if the fused approaches are complementary not fully competitive. We propose in this paper the fusion of two projection based face recognition algorithms: singular value decomposition (SVD) using the left and right singular vectors of the face image as a face feature stored in a matrix and regularized Linear Discriminant Analysis in DCT domain (DCT-RLDA) which is known by its computational efficiency in addition to discrimination power. Experiments conducted on the ORL database indicate that the application of the Min-Max, Z-score score normalization schemes followed by a simple fusion strategies (simple sum, weighted sum, append) confirm the benefits of the proposed approach in terms of identification rate and processing time.
{"title":"Score Fusion of SVD and DCT-RLDA for Face Recognition","authors":"Messaoud Bengherabi, L. Mezai, F. Harizi, A. Guessoum, M. Cheriet","doi":"10.1109/IPTA.2008.4743776","DOIUrl":"https://doi.org/10.1109/IPTA.2008.4743776","url":null,"abstract":"Although information fusion in unimodal or multimodal biometric systems can be performed at various levels, integration of the matching score level is the most common approach. Starting from the fact; that the fusion will be efficient if and only if the fused approaches are complementary not fully competitive. We propose in this paper the fusion of two projection based face recognition algorithms: singular value decomposition (SVD) using the left and right singular vectors of the face image as a face feature stored in a matrix and regularized Linear Discriminant Analysis in DCT domain (DCT-RLDA) which is known by its computational efficiency in addition to discrimination power. Experiments conducted on the ORL database indicate that the application of the Min-Max, Z-score score normalization schemes followed by a simple fusion strategies (simple sum, weighted sum, append) confirm the benefits of the proposed approach in terms of identification rate and processing time.","PeriodicalId":384072,"journal":{"name":"2008 First Workshops on Image Processing Theory, Tools and Applications","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133226786","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-11-01DOI: 10.1109/IPTA.2008.4743782
P. Rajavel, R. Aravind
This paper presents the multiscale directional decomposition based on the directional frequency information of an image. In general, the pixel values of an image predominantly changes in only a few directions, based on this fact, the directional decomposition is achieved. Two different approaches are used for decomposition namely direction dependent filter bank (DDFB) and multiscale directional Gaussian filter (MDGF). DDFB use the Laplacian pyramid for multiscale decomposition followed by DFB for directional decomposition. MDGF use the Laplacian pyramid for multiscale decomposition followed by directional Gaussian filters for directional decomposition. The number of DDFB subbands at nth stage is 3(2n-2) with a redundancy factor of 4/3. The number of MDFG subbands at nth stage is m(2n-2). This directional dependent decomposition is used for edge detection and results show the better performance compared to several edge detection techniques in the presence of noise.
{"title":"Direction Dependent Decomposition and Edge Detection","authors":"P. Rajavel, R. Aravind","doi":"10.1109/IPTA.2008.4743782","DOIUrl":"https://doi.org/10.1109/IPTA.2008.4743782","url":null,"abstract":"This paper presents the multiscale directional decomposition based on the directional frequency information of an image. In general, the pixel values of an image predominantly changes in only a few directions, based on this fact, the directional decomposition is achieved. Two different approaches are used for decomposition namely direction dependent filter bank (DDFB) and multiscale directional Gaussian filter (MDGF). DDFB use the Laplacian pyramid for multiscale decomposition followed by DFB for directional decomposition. MDGF use the Laplacian pyramid for multiscale decomposition followed by directional Gaussian filters for directional decomposition. The number of DDFB subbands at nth stage is 3(2n-2) with a redundancy factor of 4/3. The number of MDFG subbands at nth stage is m(2n-2). This directional dependent decomposition is used for edge detection and results show the better performance compared to several edge detection techniques in the presence of noise.","PeriodicalId":384072,"journal":{"name":"2008 First Workshops on Image Processing Theory, Tools and Applications","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125487430","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-11-01DOI: 10.1109/IPTA.2008.4743733
Imane M. Zendjebil, F. Ababsa, Jean-Yves Didier, M. Mallem
Outdoor Augmented Reality applications often combine heterogeneous sensors to recover 3D localization (position and orientation) in large environments. Indeed, accurate localization is critical to register virtual augmentations over a real scene. This paper describes a localization system composed of two parts: an Aid-localization subsystem and a Vision subsystem. The Aidlocalization subsystem, composed of GPS and inertial sensor, has two functionalities: to initialize the visual tracking and to estimate the user's position and orientation when visual tracking fails. The Vision subsystem which represents the main block, allows continuously estimating the user's position and orientation using point-based visual tracking.
{"title":"Hybrid Localization System for Mobile Outdoor Augmented Reality Applications","authors":"Imane M. Zendjebil, F. Ababsa, Jean-Yves Didier, M. Mallem","doi":"10.1109/IPTA.2008.4743733","DOIUrl":"https://doi.org/10.1109/IPTA.2008.4743733","url":null,"abstract":"Outdoor Augmented Reality applications often combine heterogeneous sensors to recover 3D localization (position and orientation) in large environments. Indeed, accurate localization is critical to register virtual augmentations over a real scene. This paper describes a localization system composed of two parts: an Aid-localization subsystem and a Vision subsystem. The Aidlocalization subsystem, composed of GPS and inertial sensor, has two functionalities: to initialize the visual tracking and to estimate the user's position and orientation when visual tracking fails. The Vision subsystem which represents the main block, allows continuously estimating the user's position and orientation using point-based visual tracking.","PeriodicalId":384072,"journal":{"name":"2008 First Workshops on Image Processing Theory, Tools and Applications","volume":"65 S9","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120844111","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-11-01DOI: 10.1109/IPTA.2008.4743736
A. Chaari, S. Lelandais, M. Ahmed
We propose in this paper a search approach which aim to improve identification in biometric databases. We work with face images and we develop appearance-based Eigenfaces and Fisherfaces methods to generate holistic and discriminant features and attributes. These features, which describe faces, are often used to establish the required identity in a classical identification process. In this work we introduce a clustering process upstream the identification process which divides faces into partitions according to their features similarities. Indeed, we aim to split biometric databases into partitions in order to simplify the recognition task within these databases. This paper describes the proposed clustering approach, the Eigenfaces and Fisherfaces representation methods and preliminary clustering results on the XM2VTS data corpus.
{"title":"A New Clustering Approach for Face Identification","authors":"A. Chaari, S. Lelandais, M. Ahmed","doi":"10.1109/IPTA.2008.4743736","DOIUrl":"https://doi.org/10.1109/IPTA.2008.4743736","url":null,"abstract":"We propose in this paper a search approach which aim to improve identification in biometric databases. We work with face images and we develop appearance-based Eigenfaces and Fisherfaces methods to generate holistic and discriminant features and attributes. These features, which describe faces, are often used to establish the required identity in a classical identification process. In this work we introduce a clustering process upstream the identification process which divides faces into partitions according to their features similarities. Indeed, we aim to split biometric databases into partitions in order to simplify the recognition task within these databases. This paper describes the proposed clustering approach, the Eigenfaces and Fisherfaces representation methods and preliminary clustering results on the XM2VTS data corpus.","PeriodicalId":384072,"journal":{"name":"2008 First Workshops on Image Processing Theory, Tools and Applications","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116963138","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-11-01DOI: 10.1109/IPTA.2008.4743755
Y. Ben Jemaa, S. Khanfir
In this paper we present a biometric system of face detection and recognition in color images. The face detection technique is based on skin color information. A new algorithm is proposed in order to detect automatically face features (eyes, mouth and nose) and extract their correspondent geometrical points. These fiducial points are described by sets of wavelet components called "jets" which are used for recognition. To achieve the face recognition, we propose two architectures of neural networks and we compare their performances. We also, compare the two types of features used for recognition: geometric distances and Gabor coefficients which can be used either independently or jointly. This comparison shows that Gabor coefficients are more powerful than geometric distances. We show with experimental results how the importance recognition ratio makes our system an effective tool for automatic face detection and recognition.
{"title":"Automatic Gabor Features Extraction for Face Recognition using Neural Networks","authors":"Y. Ben Jemaa, S. Khanfir","doi":"10.1109/IPTA.2008.4743755","DOIUrl":"https://doi.org/10.1109/IPTA.2008.4743755","url":null,"abstract":"In this paper we present a biometric system of face detection and recognition in color images. The face detection technique is based on skin color information. A new algorithm is proposed in order to detect automatically face features (eyes, mouth and nose) and extract their correspondent geometrical points. These fiducial points are described by sets of wavelet components called \"jets\" which are used for recognition. To achieve the face recognition, we propose two architectures of neural networks and we compare their performances. We also, compare the two types of features used for recognition: geometric distances and Gabor coefficients which can be used either independently or jointly. This comparison shows that Gabor coefficients are more powerful than geometric distances. We show with experimental results how the importance recognition ratio makes our system an effective tool for automatic face detection and recognition.","PeriodicalId":384072,"journal":{"name":"2008 First Workshops on Image Processing Theory, Tools and Applications","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127807332","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-11-01DOI: 10.1109/IPTA.2008.4743752
Heyfa Ammar, A. Benazza-Benyahia
In this paper, we are interested in detecting blotches in old movies. The novelty of our approach is twofold. Firstly, we study the pertinency of operating the detection in the wavelet transform domain. Secondly, we propose to resort to statistical outlier test in order to localize the underlying artifacts.
{"title":"Wavelet-based blotch detection in old movies","authors":"Heyfa Ammar, A. Benazza-Benyahia","doi":"10.1109/IPTA.2008.4743752","DOIUrl":"https://doi.org/10.1109/IPTA.2008.4743752","url":null,"abstract":"In this paper, we are interested in detecting blotches in old movies. The novelty of our approach is twofold. Firstly, we study the pertinency of operating the detection in the wavelet transform domain. Secondly, we propose to resort to statistical outlier test in order to localize the underlying artifacts.","PeriodicalId":384072,"journal":{"name":"2008 First Workshops on Image Processing Theory, Tools and Applications","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127687082","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-11-01DOI: 10.1109/IPTA.2008.4743788
Myung-Ho Ju, Hang-Bong Kang
This paper presents a novel method to estimate the pose of human faces from insufficient video data. We represent each pose of a person's face as a connected low-dimensional appearance manifolds which are approximated by affine plane. To construct affine planes, we first sample exemplars from video data and cluster exemplars into each pose. From exemplars, the affine plane is constructed using PCA. However, the sampled exemplars in each specific pose are often not enough for computing the affine plane. This limits the performance of pose estimation. To overcome it, we propose a new sample generation method in constructing pose manifold for on-line face manifold learning. The proposed method was evaluated under several real situations and promising results were obtained.
{"title":"Robust Face Pose Estimation from Insufficient Data","authors":"Myung-Ho Ju, Hang-Bong Kang","doi":"10.1109/IPTA.2008.4743788","DOIUrl":"https://doi.org/10.1109/IPTA.2008.4743788","url":null,"abstract":"This paper presents a novel method to estimate the pose of human faces from insufficient video data. We represent each pose of a person's face as a connected low-dimensional appearance manifolds which are approximated by affine plane. To construct affine planes, we first sample exemplars from video data and cluster exemplars into each pose. From exemplars, the affine plane is constructed using PCA. However, the sampled exemplars in each specific pose are often not enough for computing the affine plane. This limits the performance of pose estimation. To overcome it, we propose a new sample generation method in constructing pose manifold for on-line face manifold learning. The proposed method was evaluated under several real situations and promising results were obtained.","PeriodicalId":384072,"journal":{"name":"2008 First Workshops on Image Processing Theory, Tools and Applications","volume":"255 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133427996","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-11-01DOI: 10.1109/IPTA.2008.4743771
C. Beumier
This paper presents building verification in the context of change detection for database update. Disparity values between the left and right images of a stereo couple are estimated from grey level correlation in the vicinity of edge pixels and harmonized along contours for better coherence and accuracy. This estimation is very fast as only edge pixels are considered and accurate thanks to grey level and geometrical constraints. Building evidence is derived from the presence of edge points with high disparity near or inside building polygons of the database. Verification results compared favourably with a previous work based on straight line and shadow detection. The system intends to assist geographers in the process of topographic database update for the building layer. It also aims at delivering useful by-products like building elevation (Z coordinate) and Digital Terrain Model (DTM).
{"title":"Building Verification from Disparity of Contour Points","authors":"C. Beumier","doi":"10.1109/IPTA.2008.4743771","DOIUrl":"https://doi.org/10.1109/IPTA.2008.4743771","url":null,"abstract":"This paper presents building verification in the context of change detection for database update. Disparity values between the left and right images of a stereo couple are estimated from grey level correlation in the vicinity of edge pixels and harmonized along contours for better coherence and accuracy. This estimation is very fast as only edge pixels are considered and accurate thanks to grey level and geometrical constraints. Building evidence is derived from the presence of edge points with high disparity near or inside building polygons of the database. Verification results compared favourably with a previous work based on straight line and shadow detection. The system intends to assist geographers in the process of topographic database update for the building layer. It also aims at delivering useful by-products like building elevation (Z coordinate) and Digital Terrain Model (DTM).","PeriodicalId":384072,"journal":{"name":"2008 First Workshops on Image Processing Theory, Tools and Applications","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131425193","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-11-01DOI: 10.1109/IPTA.2008.4743749
Wei Ren, M. Singh, S. Singh
In this paper, we propose a novel spatio-temporal video retrieval model to extract spatio-temporal attributes for semantic video category classification using high-level reasoning of video objects and scenes. We also explore the semantic logical inference learning of video attributes based on interpreting camera movements and object spatial constraints, as well the views on temporal continuity of video. We have used Minerva international video benchmark for the analysis of our algorithm.
{"title":"A Novel Semantic Video Classification Model","authors":"Wei Ren, M. Singh, S. Singh","doi":"10.1109/IPTA.2008.4743749","DOIUrl":"https://doi.org/10.1109/IPTA.2008.4743749","url":null,"abstract":"In this paper, we propose a novel spatio-temporal video retrieval model to extract spatio-temporal attributes for semantic video category classification using high-level reasoning of video objects and scenes. We also explore the semantic logical inference learning of video attributes based on interpreting camera movements and object spatial constraints, as well the views on temporal continuity of video. We have used Minerva international video benchmark for the analysis of our algorithm.","PeriodicalId":384072,"journal":{"name":"2008 First Workshops on Image Processing Theory, Tools and Applications","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132864817","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-11-01DOI: 10.1109/IPTA.2008.4743766
Souhila Guerfi, J. Gambotto, S. Lelandais
In this paper we propose a new geometric method for 3D face recognition based on anthropometric measurements. 3D facial feature points are measured by stereovision and are used to construct a 3D signature containing distances, indices and angles between these points, in order to discriminate individuals. This approach is evaluated on the new IV2 database and results are presented an discussed.
{"title":"3D Anthropometric Signatures for Face Verification","authors":"Souhila Guerfi, J. Gambotto, S. Lelandais","doi":"10.1109/IPTA.2008.4743766","DOIUrl":"https://doi.org/10.1109/IPTA.2008.4743766","url":null,"abstract":"In this paper we propose a new geometric method for 3D face recognition based on anthropometric measurements. 3D facial feature points are measured by stereovision and are used to construct a 3D signature containing distances, indices and angles between these points, in order to discriminate individuals. This approach is evaluated on the new IV2 database and results are presented an discussed.","PeriodicalId":384072,"journal":{"name":"2008 First Workshops on Image Processing Theory, Tools and Applications","volume":"73 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133953683","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：