Skip to main content

Fahimeh Alaei

Southern Cross University, Faculty of Science and Engineering, Faculty Member

Followers

21

Following

2

Co-authors

2

Public Views

Ph.D.

less

Southern Cross University

Università Campus Bio-Medico di Roma

Mohd Dilshad Ansari

Interests

Uploads

Papers by Fahimeh Alaei

Fast local binary pattern: Application to document image retrieval

2017 International Conference on Image and Vision Computing New Zealand (IVCNZ), 2017

The volume of digitised documents is increasing every day. Thus, designing a fast document image ... more The volume of digitised documents is increasing every day. Thus, designing a fast document image retrieval method for the large volume of document images, especially when the document images are also large in size, is of high demand. As feature extraction is one of the important steps in every document image retrieval system, a feature extraction technique with a low computing time and small feature number has a direct effect on the speed of the retrieval system. In this paper, we propose a non-parametric texture feature extraction method based on summarising the local grey-level structure of the image. To extract the proposed features, the input image is, at first, divided into a set of overlapping patches of equal size. The peripheral pixels of the centre pixel in a patch are used to extract two sets of patterns. The patterns are derived from the vertical & horizontal, and diagonal & off-diagonal pixels of the patch, separately. From each set of pixels, 15 (=2^4-1) different local binary patterns are extracted in our proposed feature extraction method. Two histograms of the local binary patterns obtained from these two sets of pixels are then created and concatenated to obtain 30 features called fast local binary patterns (F-LBP). To evaluate the efficiency of the proposed feature extraction method, MTDB and ITESOFT databases were considered for experimentation. The proposed F-LBP provided promising results with lower computing time as well as smaller memory space consumption.

Review of age and gender detection methods based on handwriting analysis

by Fahimeh Alaei and Alireza Alaei

Neural Computing and Applications, 2023

Handwriting recognition and analysis has been an active area of research in the last two decades.... more Handwriting recognition and analysis has been an active area of research in the last two decades. Handwriting analysis is being studied in various fields of science, such as graphology, neurology, psychology, and computer science. Furthermore, automated handwriting analysis has several applications, including forensic, security, medical, and disease prediction. This paper presents the most recent handwriting analysis techniques and advancements available in the literature for age and gender classification/detection. Different steps, including feature extraction and classification, frequently used in the literature for age and gender detection, are discussed, and the presented works are classified according to the applied feature extraction and classification methods. The online and offline benchmark databases are also reviewed. We used a text mining technique to perform a quantitative content analysis of the presented research and better understand the cooccurrence network diagrams of age and gender classification/detection. This study is a valuable resource that provides new research directions to students and researchers interested in this field for further research and investigation.

A Comparison of Demographic Attributes Detection from Handwriting based on Traditional and Deep Learning Methods

by Fahimeh Alaei and Alireza Alaei

Document Analysis and Recognition – ICDAR 2023 Workshops, 2023

Analyzing handwritten documents and detecting demographic attributes of writers from handwritten ... more Analyzing handwritten documents and detecting demographic attributes of writers from handwritten samples has received enormous attention from various fields of research, including psychology, computer science and artificial intelligence. Auto-matic detection of age, gender, handedness, nationality, and qualification of writers based on handwritten documents has several real-world applications, such as forensics and psychology. This paper proposes two simple but effective methods to detect the demographic information of writers from offline handwritten document images. The proposed methods are based on traditional and deep learning approaches. In the tradi-tional machine learning method, the Rank Transform feature extraction method is used for measuring the intensity in handwriting images. The extracted handcrafted features are then fed into Support Vector Machine based classifiers to predict the demographical attributes of writers. In the deep learning method, a Convolutional Neural Network model based on the ResNet architecture with a fully connected layer, followed by a softmax layer is used to provide probability scores to facilitate demographic infor-mation detection. To evaluate the proposed methods and compare the results with the results in the literature, a comprehensive set of experiments was conducted on a fre-quently used benchmark database, KHATT. Both methods performed relatively well in predicting different demographic attributes. However, considering the settings in our experiments, the results obtained from the traditional model indicated better demo-graphic detection compared to the deep learning models in all the tasks.

Document Image Retrieval Based on Texture Features: A Recognition-Free Approach

2016 International Conference on Digital Image Computing: Techniques and Applications (DICTA), 2016

The tendency of current technology is towards a paperless world. Due to the rapid increase of dig... more The tendency of current technology is towards a paperless world. Due to the rapid increase of digitized documents, providing a fast and easy method for retrieval is in high demand. The aim of this paper is to examine the effectiveness of texture features for document image retrieval. Thus, segmentation-free document image retrieval using a binary texture method is proposed. In the proposed approach, local features are extracted, local grey-level structures are summarised, and their distribution is characterised using global features. The assumption is that texture properties in the text regions and non-text regions of the document images are different. This assumption is used to rank the available document images and retrieve only those, which have greatest visual similarity to a given query. The under-sampled image and sub-images of the original image are further considered to improve the retrieval results, which are up to 76.0% in the first ranking and 96.2% in the Top-10 ranking. The Media Team Oulu Document Database, which is a heterogeneous database that offers a great variety of page layouts and contents, is used for experimentation.

Texture Feature-based Document Image Retrieval

Storing and manipulating documents in digital form to contribute to a paperless society has been ... more Storing and manipulating documents in digital form to contribute to a paperless society has been the propensity of emerging technology. There has been notable growth in the variety and quantity of digitised documents, which have often been scanned/photographed and archived as images without any labelling or sufficient index information. The growth of these kinds of document images will undoubtedly continue with new technology. To provide an effective way for retrieving and organizing these document images, many techniques have been implemented in the literature. However, designing automation systems to accurately retrieve document images from archives remains a challenging problem. Finding discriminative and effective features is the fundamental task for developing an efficient retrieval system. An overview of the literature reveals that research on document image retrieval using texture-based features has not yet been broadly investigated. Texture features are suitable for large vo...

Gender Detection Based on Spatial Pyramid Matching

Document Analysis and Recognition – ICDAR 2021

Gender Detection Based on Spatial Pyramid Matching

by Fahimeh Alaei and Alireza Alaei

Document Analysis and Recognition – ICDAR 2021

Document Image Retrieval Based on Visual Saliency Maps

by Fahimeh Alaei and Alireza Alaei

There has been a massive growth in the production of various unstructured, complex and multilingu... more There has been a massive growth in the production of various unstructured, complex and multilingual digitised documents in recent years. Storing and manipulating such digitised documents towards a paperless society has been the objective of emerging technology. As the human visual system can easily distinguish the global summary of images, extracting features based on human attention from images is desirable to achieve more accurate document image retrieval results. Thus, in this research work, an appearance-based document image retrieval system using image saliency maps depending on human visual attention is proposed. The saliency map obtained from the input document image is used to generate a weighted document image. Features are then extracted from the weighted document images using the Gist operator. Then, locality-sensitive hashing is considered to compute similarity distances between a query and the document images in the knowledge-based database. To evaluate the performance of the proposed document image retrieval system MTDB, ITESOFT, and CLEF-IP datasets of document images were used for experimentation. The proposed document image retrieval system provided promising retrieval results compared to the results reported in the literature.

Evaluation of Gist Operator for Document Image Retrieval

—As digitised documents normally contain a large variety of structures, a page segmentation-and l... more —As digitised documents normally contain a large variety of structures, a page segmentation-and layout-free method for document image retrieval is preferable. In this research work, therefore, wavelet transform as a transform-based approach is initially used to provide different under-sampled images from the original image. Then, Gist operator, as a feature extraction technique, is employed to extract a set of global features from the original image as well as the sub-images obtained from the wavelet transform. Moreover, the column-wise variances of the values in each sub-image are computed and they are then concatenated to obtain another set of features. Considering each feature set, locality-sensitive hashing is employed to compute similarity distances between a query and the document images in the database. Finally, a classifier fusion technique using the mean function is taken into account to provide a document image retrieval result. The combination of these features and a clustering score fusion strategy provides higher document image retrieval accuracy. Two different databases of the document image are considered for experimentation. The results obtained from the experimental study are detailed and the results are encouraging.

Fast Local Binary Pattern: Application to Document Image Retrieval

— The volume of digitised documents is increasing every day. Thus, designing a fast document imag... more — The volume of digitised documents is increasing every day. Thus, designing a fast document image retrieval method for the large volume of document images, especially when the document images are also large in size, is of high demand. As feature extraction is one of the important steps in every document image retrieval system, a feature extraction technique with a low computing time and small feature number has a direct effect on the speed of the retrieval system. In this paper, we propose a non-parametric texture feature extraction method based on summarising the local grey-level structure of the image. To extract the proposed features, the input image is, at first, divided into a set of overlapping patches of equal size. The peripheral pixels of the centre pixel in a patch are used to extract two sets of patterns. The patterns are derived from the vertical & horizontal, and diagonal & off-diagonal pixels of the patch, separately. From each set of pixels, 15 different local binary patterns are extracted in our proposed feature extraction method. Two histograms of the local binary patterns are then created and concatenated to obtain 30 features called fast local binary pattern (F-LBP). To evaluate the efficiency of the proposed feature extraction method, MTDB and ITESOFT databases were considered for experimentation. The proposed F-LBP provided promising results with lower computing time as well as smaller memory space consumption compared to other variation of LBP methods.

Document Image Retrieval Based on Texture Features and Similarity Fusion

—In this paper we investigate the usefulness of two different texture features along with classif... more —In this paper we investigate the usefulness of two different texture features along with classification fusion for document image retrieval. A local binary texture method, as a statistical approach, and a wavelet analysis technique, as a transform-based approach, are used for feature extraction and two feature vectors are obtained for every document image. The similarity distances between each of the two feature vectors extracted for a given query and the feature vectors extracted from the document images in the training step are computed separately. In order to use the properties of both features, a classifier fusion technique is then employed using a weighted average fusion of distance measures obtained in relation to each feature vector. The document images are finally ranked based on the greatest visual similarity to the query obtained from the fusion similarity measures. The Media Team Document Database, which provides a great variety of page layouts and contents, is considered for evaluating the proposed method. The results obtained from the experiments demonstrate a correct document retrieval of 65.4% and 91.8% in the Top-1 and Top-10 ranked document list, respectively.

Document Image Retrieval Based on Texture Features: A Recognition-Free Approach

by Fahimeh Alaei and Alireza Alaei

—The tendency of current technology is towards a paperless world. Due to the rapid increase of di... more —The tendency of current technology is towards a paperless world. Due to the rapid increase of digitized documents, providing a fast and easy method for retrieval is in high demand. The aim of this paper is to examine the effectiveness of texture features for document image retrieval. Thus, segmentation-free document image retrieval using a binary texture method is proposed. In the proposed approach, local features are extracted, local grey-level structures are summarised, and their distribution is characterised using global features. The assumption is that texture properties in the text regions and non-text regions of the document images are different. This assumption is used to rank the available document images and retrieve only those, which have greatest visual similarity to a given query. The under-sampled image and sub-images of the original image are further considered to improve the retrieval results, which are up to 76.0% in the first ranking and 96.2% in the Top-10 ranking. The Media Team Oulu Document Database, which is a heterogeneous database that offers a great variety of page layouts and contents, is used for experimentation.

A Brief Review of Document Image Retrieval Methods: Recent Advances

—Due to the rapid increase of different digitized documents, the development of a system to autom... more —Due to the rapid increase of different digitized documents, the development of a system to automatically retrieve document images from a large collection of structured and unstructured document images is in high demand. Many techniques have been developed to provide an efficient and effective way for retrieving and organizing these document images in the literature. This paper provides an overview of the methods which have been applied for document image retrieval over recent years. It has been found that from a textual perspective, more attention has been paid to the feature extraction methods without using OCR.

Une nouvelle méthode de classification monoclasse à base de représentation symbolique : Application à la classification de documents

Reconnaissance de Formes et l'Intelligence Artificielle (RFIA'14), Jun 30, 2014

Construire un classificateur obtenant de bons résultats tout en utilisant un faible nombre d’exem... more Construire un classificateur obtenant de bons résultats tout en utilisant un faible nombre d’exemples d’apprentissage est un besoin récurrent dans le domaine de la classification d’images de documents, et en particulier pour l’entreprise pour laquelle cette étude a été réalisée. Dans ce cas-là, le choix d’utiliser un classificateur monoclasse (nécessitant uniquement des exemples positifs) représente une alterna- tive intéressante. Dans cet article, nous présentons une nouvelle méthode de classification monoclasse basée sur une représentation symbolique. Initialement, un ensemble de caractéristiques est extrait à partir des données de l’en- semble d’apprentissage. Puis, un vecteur d’intervalles is- sus de ces caractéristiques est construit pour représenter la classe. Chaque intervalle (donnée symbolique) est cal- culé à partir de la moyenne et de l’écart-type de chaque caractéristique. Pour évaluer le classificateur monoclasse proposé, nous avons utilisé un ensemble de données com- posé de 544 images de documents. Les expérimentations montrent que le classificateur monoclasse est performant lorsque le nombre d’exemples d’apprentissage est faible (=10). Il est donc utilisable dans un contexte de classifi- cation de documents, avec de meilleurs résultats que ceux obtenus par un classificateur k-ppv.

A New One-class Classification Method Based on Symbolic Representation: Application to Document Classification

2014 11th IAPR International Workshop on Document Analysis Systems (DAS), Apr 4, 2014

Abstract—Training a system using a small number of instances to obtain accurate recognition/clas... more Abstract—Training a system using a small number of instances
to obtain accurate recognition/classification is a crucial need in
document classification domain. The one-class classification is
chosen since only positive samples are available for the
training. In this paper, a new one-class classification method
based on symbolic representation method is proposed. Initially
a set of features is extracted from the training set. A set of
intervals valued symbolic feature vector is then used to
represent the class. Each interval value (symbolic data) is
computed using mean and standard deviation of the
corresponding feature values. To evaluate the proposed oneclass
classification method a dataset composed of 544
document images was used. Experiment results reveal that the
proposed one-class classification method works well even when
the number of training samples is small (≤10). Moreover, we
noted that the proposed one-class classification method is
suitable for document classification and provides better result
compared to one-class k-nearest neighbor (k-NN) classifier.

Accurate Person Recognition on Combining Signature and Fingerprint

International Journal of Machine Intelligence, Dec 9, 2011

Abstract- This paper is addressing the combination of physiological and behavioral biometrics. In... more Abstract- This paper is addressing the combination of physiological and behavioral biometrics. In this we adopted multimodal approach on combining signature and fingerprint biometrics using feature level fusion with different normalization techniques to improve the performance of multimodal system. Texture features are extracted for fingerprint, pen pressure, azimuth, and altitude features are extracted for signature. Thus, classification is done using K-NN and SVM classifiers; from experimental results of two classifiers shows that multimodal approach performs well by adopting robust feature normalization technique with suitable distance measures and kernels of classifiers.

Fast local binary pattern: Application to document image retrieval

2017 International Conference on Image and Vision Computing New Zealand (IVCNZ), 2017

The volume of digitised documents is increasing every day. Thus, designing a fast document image ... more The volume of digitised documents is increasing every day. Thus, designing a fast document image retrieval method for the large volume of document images, especially when the document images are also large in size, is of high demand. As feature extraction is one of the important steps in every document image retrieval system, a feature extraction technique with a low computing time and small feature number has a direct effect on the speed of the retrieval system. In this paper, we propose a non-parametric texture feature extraction method based on summarising the local grey-level structure of the image. To extract the proposed features, the input image is, at first, divided into a set of overlapping patches of equal size. The peripheral pixels of the centre pixel in a patch are used to extract two sets of patterns. The patterns are derived from the vertical & horizontal, and diagonal & off-diagonal pixels of the patch, separately. From each set of pixels, 15 (=2^4-1) different local binary patterns are extracted in our proposed feature extraction method. Two histograms of the local binary patterns obtained from these two sets of pixels are then created and concatenated to obtain 30 features called fast local binary patterns (F-LBP). To evaluate the efficiency of the proposed feature extraction method, MTDB and ITESOFT databases were considered for experimentation. The proposed F-LBP provided promising results with lower computing time as well as smaller memory space consumption.

Review of age and gender detection methods based on handwriting analysis

by Fahimeh Alaei and Alireza Alaei

Neural Computing and Applications, 2023

Handwriting recognition and analysis has been an active area of research in the last two decades.... more Handwriting recognition and analysis has been an active area of research in the last two decades. Handwriting analysis is being studied in various fields of science, such as graphology, neurology, psychology, and computer science. Furthermore, automated handwriting analysis has several applications, including forensic, security, medical, and disease prediction. This paper presents the most recent handwriting analysis techniques and advancements available in the literature for age and gender classification/detection. Different steps, including feature extraction and classification, frequently used in the literature for age and gender detection, are discussed, and the presented works are classified according to the applied feature extraction and classification methods. The online and offline benchmark databases are also reviewed. We used a text mining technique to perform a quantitative content analysis of the presented research and better understand the cooccurrence network diagrams of age and gender classification/detection. This study is a valuable resource that provides new research directions to students and researchers interested in this field for further research and investigation.

A Comparison of Demographic Attributes Detection from Handwriting based on Traditional and Deep Learning Methods

by Fahimeh Alaei and Alireza Alaei

Document Analysis and Recognition – ICDAR 2023 Workshops, 2023

Analyzing handwritten documents and detecting demographic attributes of writers from handwritten ... more Analyzing handwritten documents and detecting demographic attributes of writers from handwritten samples has received enormous attention from various fields of research, including psychology, computer science and artificial intelligence. Auto-matic detection of age, gender, handedness, nationality, and qualification of writers based on handwritten documents has several real-world applications, such as forensics and psychology. This paper proposes two simple but effective methods to detect the demographic information of writers from offline handwritten document images. The proposed methods are based on traditional and deep learning approaches. In the tradi-tional machine learning method, the Rank Transform feature extraction method is used for measuring the intensity in handwriting images. The extracted handcrafted features are then fed into Support Vector Machine based classifiers to predict the demographical attributes of writers. In the deep learning method, a Convolutional Neural Network model based on the ResNet architecture with a fully connected layer, followed by a softmax layer is used to provide probability scores to facilitate demographic infor-mation detection. To evaluate the proposed methods and compare the results with the results in the literature, a comprehensive set of experiments was conducted on a fre-quently used benchmark database, KHATT. Both methods performed relatively well in predicting different demographic attributes. However, considering the settings in our experiments, the results obtained from the traditional model indicated better demo-graphic detection compared to the deep learning models in all the tasks.

Document Image Retrieval Based on Texture Features: A Recognition-Free Approach

2016 International Conference on Digital Image Computing: Techniques and Applications (DICTA), 2016

The tendency of current technology is towards a paperless world. Due to the rapid increase of dig... more The tendency of current technology is towards a paperless world. Due to the rapid increase of digitized documents, providing a fast and easy method for retrieval is in high demand. The aim of this paper is to examine the effectiveness of texture features for document image retrieval. Thus, segmentation-free document image retrieval using a binary texture method is proposed. In the proposed approach, local features are extracted, local grey-level structures are summarised, and their distribution is characterised using global features. The assumption is that texture properties in the text regions and non-text regions of the document images are different. This assumption is used to rank the available document images and retrieve only those, which have greatest visual similarity to a given query. The under-sampled image and sub-images of the original image are further considered to improve the retrieval results, which are up to 76.0% in the first ranking and 96.2% in the Top-10 ranking. The Media Team Oulu Document Database, which is a heterogeneous database that offers a great variety of page layouts and contents, is used for experimentation.

Texture Feature-based Document Image Retrieval

Storing and manipulating documents in digital form to contribute to a paperless society has been ... more Storing and manipulating documents in digital form to contribute to a paperless society has been the propensity of emerging technology. There has been notable growth in the variety and quantity of digitised documents, which have often been scanned/photographed and archived as images without any labelling or sufficient index information. The growth of these kinds of document images will undoubtedly continue with new technology. To provide an effective way for retrieving and organizing these document images, many techniques have been implemented in the literature. However, designing automation systems to accurately retrieve document images from archives remains a challenging problem. Finding discriminative and effective features is the fundamental task for developing an efficient retrieval system. An overview of the literature reveals that research on document image retrieval using texture-based features has not yet been broadly investigated. Texture features are suitable for large vo...

Gender Detection Based on Spatial Pyramid Matching

Document Analysis and Recognition – ICDAR 2021

Gender Detection Based on Spatial Pyramid Matching

by Fahimeh Alaei and Alireza Alaei

Document Analysis and Recognition – ICDAR 2021

Document Image Retrieval Based on Visual Saliency Maps

by Fahimeh Alaei and Alireza Alaei

There has been a massive growth in the production of various unstructured, complex and multilingu... more There has been a massive growth in the production of various unstructured, complex and multilingual digitised documents in recent years. Storing and manipulating such digitised documents towards a paperless society has been the objective of emerging technology. As the human visual system can easily distinguish the global summary of images, extracting features based on human attention from images is desirable to achieve more accurate document image retrieval results. Thus, in this research work, an appearance-based document image retrieval system using image saliency maps depending on human visual attention is proposed. The saliency map obtained from the input document image is used to generate a weighted document image. Features are then extracted from the weighted document images using the Gist operator. Then, locality-sensitive hashing is considered to compute similarity distances between a query and the document images in the knowledge-based database. To evaluate the performance of the proposed document image retrieval system MTDB, ITESOFT, and CLEF-IP datasets of document images were used for experimentation. The proposed document image retrieval system provided promising retrieval results compared to the results reported in the literature.

Evaluation of Gist Operator for Document Image Retrieval

—As digitised documents normally contain a large variety of structures, a page segmentation-and l... more —As digitised documents normally contain a large variety of structures, a page segmentation-and layout-free method for document image retrieval is preferable. In this research work, therefore, wavelet transform as a transform-based approach is initially used to provide different under-sampled images from the original image. Then, Gist operator, as a feature extraction technique, is employed to extract a set of global features from the original image as well as the sub-images obtained from the wavelet transform. Moreover, the column-wise variances of the values in each sub-image are computed and they are then concatenated to obtain another set of features. Considering each feature set, locality-sensitive hashing is employed to compute similarity distances between a query and the document images in the database. Finally, a classifier fusion technique using the mean function is taken into account to provide a document image retrieval result. The combination of these features and a clustering score fusion strategy provides higher document image retrieval accuracy. Two different databases of the document image are considered for experimentation. The results obtained from the experimental study are detailed and the results are encouraging.

Fast Local Binary Pattern: Application to Document Image Retrieval

— The volume of digitised documents is increasing every day. Thus, designing a fast document imag... more — The volume of digitised documents is increasing every day. Thus, designing a fast document image retrieval method for the large volume of document images, especially when the document images are also large in size, is of high demand. As feature extraction is one of the important steps in every document image retrieval system, a feature extraction technique with a low computing time and small feature number has a direct effect on the speed of the retrieval system. In this paper, we propose a non-parametric texture feature extraction method based on summarising the local grey-level structure of the image. To extract the proposed features, the input image is, at first, divided into a set of overlapping patches of equal size. The peripheral pixels of the centre pixel in a patch are used to extract two sets of patterns. The patterns are derived from the vertical & horizontal, and diagonal & off-diagonal pixels of the patch, separately. From each set of pixels, 15 different local binary patterns are extracted in our proposed feature extraction method. Two histograms of the local binary patterns are then created and concatenated to obtain 30 features called fast local binary pattern (F-LBP). To evaluate the efficiency of the proposed feature extraction method, MTDB and ITESOFT databases were considered for experimentation. The proposed F-LBP provided promising results with lower computing time as well as smaller memory space consumption compared to other variation of LBP methods.

Document Image Retrieval Based on Texture Features and Similarity Fusion

—In this paper we investigate the usefulness of two different texture features along with classif... more —In this paper we investigate the usefulness of two different texture features along with classification fusion for document image retrieval. A local binary texture method, as a statistical approach, and a wavelet analysis technique, as a transform-based approach, are used for feature extraction and two feature vectors are obtained for every document image. The similarity distances between each of the two feature vectors extracted for a given query and the feature vectors extracted from the document images in the training step are computed separately. In order to use the properties of both features, a classifier fusion technique is then employed using a weighted average fusion of distance measures obtained in relation to each feature vector. The document images are finally ranked based on the greatest visual similarity to the query obtained from the fusion similarity measures. The Media Team Document Database, which provides a great variety of page layouts and contents, is considered for evaluating the proposed method. The results obtained from the experiments demonstrate a correct document retrieval of 65.4% and 91.8% in the Top-1 and Top-10 ranked document list, respectively.

Document Image Retrieval Based on Texture Features: A Recognition-Free Approach

by Fahimeh Alaei and Alireza Alaei

—The tendency of current technology is towards a paperless world. Due to the rapid increase of di... more —The tendency of current technology is towards a paperless world. Due to the rapid increase of digitized documents, providing a fast and easy method for retrieval is in high demand. The aim of this paper is to examine the effectiveness of texture features for document image retrieval. Thus, segmentation-free document image retrieval using a binary texture method is proposed. In the proposed approach, local features are extracted, local grey-level structures are summarised, and their distribution is characterised using global features. The assumption is that texture properties in the text regions and non-text regions of the document images are different. This assumption is used to rank the available document images and retrieve only those, which have greatest visual similarity to a given query. The under-sampled image and sub-images of the original image are further considered to improve the retrieval results, which are up to 76.0% in the first ranking and 96.2% in the Top-10 ranking. The Media Team Oulu Document Database, which is a heterogeneous database that offers a great variety of page layouts and contents, is used for experimentation.

A Brief Review of Document Image Retrieval Methods: Recent Advances

—Due to the rapid increase of different digitized documents, the development of a system to autom... more —Due to the rapid increase of different digitized documents, the development of a system to automatically retrieve document images from a large collection of structured and unstructured document images is in high demand. Many techniques have been developed to provide an efficient and effective way for retrieving and organizing these document images in the literature. This paper provides an overview of the methods which have been applied for document image retrieval over recent years. It has been found that from a textual perspective, more attention has been paid to the feature extraction methods without using OCR.

Une nouvelle méthode de classification monoclasse à base de représentation symbolique : Application à la classification de documents

Reconnaissance de Formes et l'Intelligence Artificielle (RFIA'14), Jun 30, 2014

Construire un classificateur obtenant de bons résultats tout en utilisant un faible nombre d’exem... more Construire un classificateur obtenant de bons résultats tout en utilisant un faible nombre d’exemples d’apprentissage est un besoin récurrent dans le domaine de la classification d’images de documents, et en particulier pour l’entreprise pour laquelle cette étude a été réalisée. Dans ce cas-là, le choix d’utiliser un classificateur monoclasse (nécessitant uniquement des exemples positifs) représente une alterna- tive intéressante. Dans cet article, nous présentons une nouvelle méthode de classification monoclasse basée sur une représentation symbolique. Initialement, un ensemble de caractéristiques est extrait à partir des données de l’en- semble d’apprentissage. Puis, un vecteur d’intervalles is- sus de ces caractéristiques est construit pour représenter la classe. Chaque intervalle (donnée symbolique) est cal- culé à partir de la moyenne et de l’écart-type de chaque caractéristique. Pour évaluer le classificateur monoclasse proposé, nous avons utilisé un ensemble de données com- posé de 544 images de documents. Les expérimentations montrent que le classificateur monoclasse est performant lorsque le nombre d’exemples d’apprentissage est faible (=10). Il est donc utilisable dans un contexte de classifi- cation de documents, avec de meilleurs résultats que ceux obtenus par un classificateur k-ppv.

A New One-class Classification Method Based on Symbolic Representation: Application to Document Classification

2014 11th IAPR International Workshop on Document Analysis Systems (DAS), Apr 4, 2014

Abstract—Training a system using a small number of instances to obtain accurate recognition/clas... more Abstract—Training a system using a small number of instances
to obtain accurate recognition/classification is a crucial need in
document classification domain. The one-class classification is
chosen since only positive samples are available for the
training. In this paper, a new one-class classification method
based on symbolic representation method is proposed. Initially
a set of features is extracted from the training set. A set of
intervals valued symbolic feature vector is then used to
represent the class. Each interval value (symbolic data) is
computed using mean and standard deviation of the
corresponding feature values. To evaluate the proposed oneclass
classification method a dataset composed of 544
document images was used. Experiment results reveal that the
proposed one-class classification method works well even when
the number of training samples is small (≤10). Moreover, we
noted that the proposed one-class classification method is
suitable for document classification and provides better result
compared to one-class k-nearest neighbor (k-NN) classifier.

Accurate Person Recognition on Combining Signature and Fingerprint

International Journal of Machine Intelligence, Dec 9, 2011

Abstract- This paper is addressing the combination of physiological and behavioral biometrics. In... more Abstract- This paper is addressing the combination of physiological and behavioral biometrics. In this we adopted multimodal approach on combining signature and fingerprint biometrics using feature level fusion with different normalization techniques to improve the performance of multimodal system. Texture features are extracted for fingerprint, pen pressure, azimuth, and altitude features are extracted for signature. Thus, classification is done using K-NN and SVM classifiers; from experimental results of two classifiers shows that multimodal approach performs well by adopting robust feature normalization technique with suitable distance measures and kernels of classifiers.