Year 2008

• N. Barreira, M. G. Penedo, C. Alonso-Montes, J. Rouco, "Handling topological changes in the Topological Active Volumes model ", Lecture Notes in Computer Science: Image Analysis and Recognition., 5112, 122-131, 2008.
  [Abstract] [PDF] [+Info.]
  The Topological Active Volumes (TAV) model is a 3D deformable model based on the active nets and used for segmentation and reconstruction tasks. The model implements automatic procedures, the so called topological changes, that alter the mesh structure in order to segment complex surfaces, such as pronounced curvatures or holes, and detect several objects in the scene.This work analyses previous strategies for performing topological changes in the model and proposes a new methodology that overcomes the limitations of former strategies and improves the adjustment.
• J. Novo, M. G. Penedo, J. Santos, "Optic disc segmentation by means of GA-Optimized Topological Active Nets", Lecture Notes in Computer Science: Image Analysis and Recognition., 5112, 807-816, 2008.
  [Abstract] [PDF] [+Info.]
  In this paper we propose a new approach to the optic disc segmentation process in digital retinal images by means of Topological Active Nets (TAN). This is a deformable model used for image segmentation that integrates features of region-based and edge-based segmentation techniques, being able to fit the edges of the objects and model their inner topology. The optimization of the Active Nets is performed by a genetic algorithm, with adapted or new ad hoc genetic operators to the problem. The active nets incorporate new energy terms for the optic disc segmentations, without the need of any pre-processing of the images. We present results of optic disc segmentations showing the advantages of the approach.
• E. Ares, M. G. Penedo, M. Ortega, C. Mariño, N. Barreira, F. Gómez-Ulla, "Automated three stage red lesions detection in digital color fundus images", WSEAS Transactions on Computers, 7 (4), 207-215, 2008.
  [Abstract] [PDF]
  The screening process is a very valuable method for prevention of many pathologies including Diabetic retinopathy. Tipically, a large amount of images have to be analyzed as diabetic patients have both their eyes examined at least once a year. To effectively manage all this information and the workload it produces, automatic techniques for analyzing the images are required. These techniques must be robust, sensitive and specific to be implemented in real-life screening applications. In this work an algorithm for the detection of red lesions in digital color fundus photographs is proposed. The method performs in three stages: in the first stage points candidates to be red lesions are obtained by using a set of correlation filters working in different resolutions, allowing that way the detection of a wider set of points. Then, in the second stage, a region growing segmentation process rejects the points from the prior stage whose size does not fit in the red lesion pattern. Finally, in the third stage three test are applied to the output of the second stage: a shape test to remove non-circular areas, an intensity test to remove that areas belonging to the fundus of the retina and finally a test to remove the points which fall inside the vessels (only lesions outside the vessels are considered). Evaluation was performed on a test composed on images representative of those normally found in a screening set. Moreover, comparison with manually-obtained results from clinical experts are performed, to set the accuracy and reliability of the method.
• S. Pena, M. G. Penedo, M. Ortega, J. Rouco, C. Mariño, A. Pose-Reino, M. Pena-Seijo, "Macula precise localization using digital retinal angiographies", WSEAS Transactions on Computer Research, 3 (1), 43-50, 2008.
  [Abstract] [PDF]
  The fovea is a spot located in the center of the macula, and responsible for sharp central vision. In this paper a method to detect the macula location and size is presented, as a first step towards the fovea location. Prior to the macula detection, the optic disk size and position is computed. This is performed through the combination of two stages: firstly, a clustering algorithm is used to select the regions which contain the pixels with the highest gray levels. A correlation filter is applied to these regions to compute the approximate center of the optic disk. Then, in order to extract the optic disk, a deformable model which is used. Then, following the morphological properties of the eye, the macula location and size is determined by means of a new correlation filter. Search with this filter is performed in a reduced area of interest, whose size and position is determined by means, again, of the morphological properties of the eye. The algorithm has proven to be fast and accurate in the set of test images, composed by 135 digital retinal images.
• M. Penas, M. G. Penedo, M. J. Carreira, "A neural network based framework for directional primitive extraction", Neural Processing Letters, 27 (1), 67-83, 2008.
  [Abstract] [PDF] [+Info.]
  This paper describes a computational framework for the extraction of low-level directional primitives in images. The system is divided in two stages. The first one consists in the extraction of the low level directional primitives, through the Gabor wavelet decomposition. The second one consists in the reduction of the high dimensionality of the Gabor decomposition results by means of auto-organised structures. The main advantages of the system introduced are two: it provides accurate and reliable information, and it produces good results on different image types without intervention of the final user. These advantages will be demonstrated by comparing our system with a classical edge detector.
• L. Espona, M. J. Carreira, M. G. Penedo, M. Ortega, "Comparison of Pixel and Subpixel Retinal Vessel Tree Segmentation using a Deformable Contour Model", LNCS, 13th Iberoamerican Congress in Pattern Recognition, CIARP, 5197, 683-690, 2008.
  [Abstract] [PDF]
  This paper presents a comparison of pixel and subpixel performance of the snake-based system designed to detect the vessel tree in eye fundus images. The automatic analysis of the retinal vessel tree facilitates the computation of the arteriovenous index, which is essential for the diagnosis and evolution of several eye diseases. A high accuracy is required to correctly assess the clinicians and it is insufficient when working at a pixel level. The developed model is inspired in the classical snake but incorporating domain specific knowledge and profits from the automatic localization of the optic disc and from the extraction of vascular tree centerlines previously developed in our research group [1]. The efficiency and accuracy of the detection of arteriovenous structures are evaluated using the publicly available DRIVE database and an equivalent system configuration for pixel and subpixel results. Results demonstrate that, although more time consuming, subpixel retinal vessel extraction is much more reliable, keeping relatively low values of computing time.
• C. Alonso-Montes, D. L. Vilarino, P. Dudek, M. G. Penedo, "Fast Retinal Vessel Tree Extraction: A Pixel Parallel Approach", International Journal of Circuit Theory and Application (Special Issue Cellular Wave Computing Architectures), 36, 641-651, 2008.
  [Abstract] [PDF] [+Info.]
  Early ocular disease diagnosis has became one of the main fields in medical research. From the image processing point of view, many strategies and algorithms have been developed to deal with the extraction of the retinal vessel tree. Although reliable and accurate results have been obtained, the main disadvantage in most of these proposals is still the high computation effort required. In this paper, a methodology to extract the retinal vessel tree has been developed and tested in a fine-grain pixel-parallel processor array. An analysis of the execution time has been made to show its capabilities regarding the computation speed. Moreover, an analysis of the accuracy using a publicly available database has been made to show the algorithm performance.
• M. Ortega, M. G. Penedo, C. Mariño, M. J. Carreira, "Similarity Metrics Analysis for Feature Point Based Retinal Authentication", Lecture Notes in Computer Science: Image Analysis and Recognition., 5112, 1023-1032, 2008.
  [Abstract] [PDF] [+Info.]
  Biometrics refer to identity verification of individuals based on some physiologic or behavioural characteristics (face, fingerprint, signature...). The typical authentication process of a person consists in extracting a biometric pattern of him/her and matching it with the stored pattern for the authorized user obtaining a similarity value between patterns. If that similarity is bigger than some threshold the authentication is accepted, otherwise is rejected. Thus, the similarity metrics determine the system ability to successfully classify authentications as authorized or unauthorized. In this work, an analysis of similarity metrics performance is presented for a biometric system in which retinal vessel feature points are used as biometric pattern. The results of the system allow to establish a confidence band for the metric threshold where no errors are obtained for training and test sets.
• C. Mariño, M. G. Penedo, J. Rouco, M. Ortega, A. Pose-Reino, "MedicalWeb Application For The Evaluation of Changes in Retinal Microcirculation", Computational Engineering in Systems Applications, 113-119, 2008.
  [Abstract] [PDF] [+Info.]
  Arterial Hypertension, most commonly referred to as ”high blood pressure”, HTN or HPN, is one of the most frequent condition among the Spanish population, with an estimated prevalence, in people older than 18 years, of 35%, with values of 40% in middle range ages and 68% in people older than 60 years. In hypertense people, lessions in the eye fundus appear at the same time than the vascular cerebral lessions. This is the reason why the evaluation of retinal microcirculation is so important in the prevention of cerebral (as in ocular) lessions. But this prevention is associated with screening processes to a large amount of patients whose eyes are examined at least once a year. To effectively manage all this information and the workload produced by it, automatic techniques for analyzing the images are required, as they are information technology systems which allow for an effective information storage and retrieval. Here such a system is presented, where automatic eye processing and storage of the obtained information is performed using a web medical application named SIRIUS. This system is currently being used by several hospitals to analyze the results of the application of a medicine to hypertense patients.

Year 2007

• J. Santos, O. Ibáñez, N. Barreira, M. G. Penedo, "Genetic-Greedy Hybrid Approach for Topological Active Nets Optimization", Lecture Notes in Computer Science: Adaptive and Natural Computing Algorithms, 4431, 202-210, 2007.
  [Abstract] [PDF]
  In this paper we propose a genetic and greedy algorithm combination for the optimization of the Topological Active Nets (TAN) model. This is a deformable model used for image segmentation that integrates features of region-based and edge-based segmentation techniques, being able to fit the edges of the objects and model their inner topology. The hybrid approach we propose can optimize the active nets through the minimization of the model energy functions and, moreover, it can provide some segmentation results unreachable by the GA method alone such as changes in the net topology.
• J. Rouco, M. Penas, M. G. Penedo, M. Ortega, C. Alonso-Montes, "Certainty Measure of Pairwise Line Segment Perceptual Relations using Fuzzy Logic", LNCS, Progress in Pattern Recognition, Image Analysis and Applications, 4756, 477-486, 2007.
  [Abstract] [PDF] [+Info.]
  Perceptual grouping is an important part of many computer vision systems. When inferring a new grouping from the primitive features there is always an uncertainty degree on this detection, that might be useful in further reasonings. In this paper, we present a fuzzy logic based system for the computation of the certainties assigned to pairwise line segment relations and introduce its application to the detection of continuity, identity, junction, L-junction, incidence, T-junction and parallelism relations. The results presented show that the proposed method might be very promising for future applications.
• M. Penas, M. J. Carreira, M. G. Penedo, N. Barreira, "Comparison of Alternative Frameworks for Directional Primitive Estraction", Pattern Recognition and Image Analysis, 17, Nº 4, 439-449, 2007.
  [Abstract] [PDF] [+Info.]
  This paper describes and compares three alternative computational frameworks for the extraction of the directional primitives present in an image. These frameworks are based on Gabor decomposition, Sobel filtering, and filtering through the first derivative of a Gaussian filter, respectively. They are divided into two stages: low-level primitive extraction and organization of low-level primitives by means of dynamical neural networks. These alternative frameworks are compared and their relative advantages and disadvantages are out-lined.
• L. Espona, M. J. Carreira, M. Ortega, M. G. Penedo, "A Snake for Retinal Vessel Segmentation", Lecture Notes in Computer Science: Pattern Recognition and Image Analysis. , 4478, 178-185, 2007.
  [Abstract] [PDF] [+Info.]
  This paper presents an innovative methodology to detect the vessel tree in retinal angiographies. The automatic analysis of retinal vessel tree facilitates the computation of the arteriovenous index, which is essential for the diagnosis of a wide range of eye diseases. We have developed a system inspired in the classical snake but incorporating domain specific knowledge, such as blood vessels topological properties. It profites mainly from the automatic localization of the optic disc and from the extraction and enhancement of the vascular tree centerlines. Encouraging results in the detection of arteriovenous structures are efficiently achieved, as shown by the systems performance evaluation on the publicy available DRIVE database.
• N. Barreira, M. G. Penedo, O. Ibáñez, J. Santos, "Automatic Topological Active Net Division in a Genetic-Greedy Hybrid Approach", Lecture Notes in Computer Science: Pattern Recognition and Image Analysis, 4478, 226-233, 2007.
  [Abstract] [PDF] [+Info.]
  In this paper we propose an automatic division procedure for the Topological Active Net model in a hybrid combination of a genetic and a greedy algorithm. This procedure allows the division of the active net in subnets with the aim of segmenting several objects in the same image. The combination of the greedy algorithm and the global search improves the results in both synthetic and real images.
• M. Penas, M. J. Carreira, M. G. Penedo, N. Barreira, "Efficient combination of the fuzzy Hough transform and the Burns segment detector", Lecture Notes in Computer Science (selected papers EUROCAST'07), 4739, 733-739, 2007.
  [Abstract] [PDF] [+Info.]
  This paper describes a computational framework for the fully automated detection of line segments in 2D digital images. The operation of the framework is divided in two stages, the low level directional primitive detection through Gabor wavelets and growing cell structures, and the segment detection through an efficient and very accurate combination of the fuzzy Hough transform and the Burns segment detector.
• C. Mariño, N. Barreira, M. G. Penedo, M. Ortega, J. L. Doncel, F. Gómez-Ulla, "Two stages optic disc segmentation in digital retinal images", WSEAS Transactions on Information Science and Applications, 4/4, 771-778, 2007.
  [Abstract] [PDF] [+Info.]
  In this work an algorithm for the automatic extraction of the optic disc in retinal digital images is presented. The developed system consists of two main parts. Firstly, a clustering algorithm is used to select the regions which contain the pixels with the highest gray levels. Since several regions can be obtained from the clustering algorithm, a method to extract the right one which contains the optic disc is needed. By using a correlation filter two goals are accomplished: the clusters which does not contain with the optic disc can be discarded, and the approximate center of the optic disc is computed by taking the location of the maximum value of the correlation. Then, in order to extract the optic disc, a deformable model which works in three stages is used. The localization of the region of interest is successfully obtained in 98.5% of the cases and the optic disc is correctly extracted in 93% of the cases.

Year 2006

• S. Pan, G. Shavit, M. Penas, D. Xu, L. Shapiro, R. Ladner, E. Riskin, W. Hol, D. Meldrum, "Automated classification of protein crystallization images using support vector machines with scale-invariant texture and Gabor features", Acta Crystallographica Section D, D62, 271-279, 2006.
  [PDF]
• J. M. Cardesín, M. G. Penedo, "The Role of Multimedia Websites in Teaching and Research", Teaching Urban History in Europe;Ed. R. Rodger and D. Menjot. Essays in Urban History; Editorial Center For Urban History, 121-131, 2006.
  [+Info.]
• L. Espona, M. J. Carreira, M. G. Penedo, "Vascular Tree Segmantation in Retinal Agiographies: A Snake Model-Approach", WSEAS Transactions on Computers, 10(5), 2285-2292, 2006.
  [PDF] [+Info.]
• M. Blanco, M. G. Penedo, N. Barreira, M. Penas, M. J. Carreira, "Localization and extraction of the Optic Disc using the Fuzzy Circular Hough Transform", Lecture Notes Artificial Intelligence: Artificial Intelligence and Soft Computing , 4029, 713-721, 2006.
  [PDF] [+Info.]
• C. Mariño, M. G. Penedo, M. Penas, M. J. Carreira, F. González, "Personal authentication using digital retinal images", Pattern Analysis and Applications, 9, 21-33, 2006.
  [Abstract] [PDF] [+Info.]
  Traditional authentication (identity verification) systems, employed to gain access to a private area in a building or to data stored in a computer, are based on something the user "has" (an authentication card, a magnetic key) or something the user "knows" (a password, an identification code). But emerging technologies allow for more reliable and comfortable for the user, authentication methods, most of them based in biometric parameters. Much work could be found in literature about biometric based authentication, using parameters like iris, voice, fingerprint, face characteristics, and others. In this work a novel authentication method is presented, and first results obtained are shown. The biometric parameter employed for the authentication is the retinal vessel tree, acquired through a retinal angiography. It has already been asserted by expert clinicians that the configuration of the retinal vessels is unique for each individual and that it does not vary in his life, so it is a very well suited identification characteristic. Before the verification process can be executed, a registration step is needed to align both the reference image and the picture to be verified. A fast and reliable registration method is used to perform that step, so that the whole authentication process takes very little time.
• M. Ortega, C. Mariño, M. G. Penedo, M. Blanco, F. González, "Personal Authentication Based on Feature Extraction and Optica Nerve Location in Digital Retinal Images", WSEAS Transactions on Computers, 5(6), 1169-1176, 2006.
  [Abstract] [PDF] [+Info.]
  Authentication methods based on biometric parameters are gaining importance over the years. But they require fast and robust image processing techniques to perform this task. In this work an efficient method for the identity verification of persons based on matching digital retinal images is introduced. The matching process works by extracting a set of feature points and registering them to measure the degree of similarity between the input images. The feature points are the ridge endings and ridge bifurcations from vessels obtained from a crease model of the retinal vessel tree. The method is developed and tested to obtain a good set of feature points. Then, pairs of sets will be matched in order to get an accurate and reliable similarity measure useful for the authentication procedure.
• O. Ibáñez, N. Barreira, J. Santos, M. G. Penedo, "Topológical Active Nets Optimization Using Genetic Algorithms", Lecture Notes in Computer Science: Image Analysis and Recognition, 4141, 272-282, 2006.
  [Abstract] [PDF] [+Info.]
  The Topological Active Net (TAN) model is a deformable model used for image segmentation. It integrates features of region–based and edge–based segmentation techniques. This way, the model is able to fit the edges of the objects and model their inner topology. The model consists of a two dimensional mesh controlled by energy functions. The minimization of these energy functions leads to the TAN adjustment. This paper presents a new approach to the energy minimization process based on genetic algorithms (GA), that defines several suitable genetic operators for the optimization task. The results of the new GA approach are compared to the results of a greedy algorithm developed for the same task.

Year 2005

• C. Mariño, M. G. Penedo, M. Penas, "Retinal Based Authentication via Distributed Web Application", Lecture Notes in Computer Science, 3643, 286-391, 2005.
  [Abstract] [PDF] [+Info.]
  Traditional authentication systems, employed to gain access to a private area in a building or to data stored in a computer, are based on something the user has (an authentication card, a magnetic key) or something the user knows (a password, an identification code). But emerging technologies allow for more reliable and comfortable for the user, authentication methods, most of them based on biometric parameters. Much work could be found in literature about biometric based authentication, using parameters like iris, voice, fingerprints, face characteristics, and others. We have developed a new methodology for personal authentication, where the biometric parameter employed for the authentication is the retinal vessel tree, acquired through a retinal angiography. It has already been asserted by expert clinicians that the configuration of the retinal vessels is unique for each individual and that it does not vary in his life, so it is a very well suited identification characteristic. In this work we will present the design and implementation stages of an application which allows for a reliable personal authentication in high security environments based on the retinal authentication method.
• N. Barreira, M. G. Penedo, "Topological Active Volumes", Advances in Intelligent Vision Systems: Methods and Applications; EURASIP Journal on Applied Signal Processing, 13(1), 1939-1947, 2005.
  [Abstract] [PDF] [+Info.]
  The Topological Active Volumes (TAV) model is a general model for 3D image segmentation. It is based on deformable models and integrates features of region-based and boundary-based segmentation techniques. Besides segmentation, it can also be used for surface reconstruction and topological analysis of the inside of detected objects. The TAV structure is flexible and allows topological changes in order to improve the adjustment to object's local characteristics, find several objects in the scene and identify and delimit holes in detected structures. This paper describes the main features of the TAV model and shows its ability to segment volumes in an automated manner.
• C. Alonso-Montes, D. L. Vilarino, M. G. Penedo, "On the Automatic 2D Retinal Vessel Extraction", Lecture Notes in Computer Science: Pattern Recognition and Image Analysis, 3687(1), 165-173, 2005.
  [Abstract] [PDF] [+Info.]
  Retinal vessel extraction has become an important task of medical image processing applications in order to diagnose ocular diseases. In this paper, a novel methodology is proposed in order to extract vessels automatically from retinal angiographies. The proposed methodology has been implemented by means of Cellular Neural Networks techniques in order to take advantage of their capabilities of massively parallel processing reducing computation time required.
• J. Rouco, N. Barreira, M. G. Penedo, J. M. Pardo, "Conversion into Three-dimensional Implicit Surface Representation from Topological Active Volumes based Segmentation", Lecture Notes in Computer Science: Pattern Recognition and Image Analysis, 3522(1), 60-68, 2005.
  [Abstract] [PDF] [+Info.]
  In the last few years, the advances in three-dimensional medical image processing have made possible operations like planning or simulation over real data. Different representations of structures or models have been proposed, being the implicit surfaces one of the most flexible models for processing. This paper introduces a new method for computing the implicit surfaces from the explicit representations of the objects segmented in three-dimensional images. This proposal is based on the approximation of the surfaces using distance functions and natural neighbor interpolation. The system has been tested over CT images of tibia and femur where the explicit representation has been extracted through a TAV model. The results obtained show the suitability of the method for a correct representation of the target objects.
• I. G. Caderno, M. G. Penedo, N. Barreira, C. Mariño, F. González, "Precise Detection and Measurement of the Retina Vascular Tree", Pattern Recognition and Image Analysis: Advances in Mathematical Theory and Applications (IAPC Nauka/Interperiodica), 15(2), 523-526, 2005.
  [Abstract] [PDF] [+Info.]
  A high accuracy measurement is usually required by ophthalmologists to examine the retina vascular tree. Many comprehensive models are available, but the results are not always satisfactory. In this paper, we describe a novel method for locating and measuring blood vessels in regions of interest from retinal images. A crease extraction paradigm is used for detection, whereas a specialized deformable model is employed for diameter estimation. The method is able to locate 90% of the structures, giving a reliability of 99% in detection and 95% in measurement.
• M. Penas, M. G. Penedo, M. J. Carreira, N. Barreira, "Comparison of Alternative Frameworks for Directional Primitive Extraction", Pattern Recognition and Image Analysis: Advances in Mathematical Theory and Applications (IAPC Nauka/Interperiodica), 15(2), 279-282, 2005.
  [Abstract] [PDF] [+Info.]
  This paper describes and compares two alternative frameworks for detecting the directional primitives present in an image. The first framework is based on Gabor wavelet decomposition and the second on the edge detector that is traditionally considered to be optimal: the first derivative of a Gaussian. Both frameworks are divided into two stages: low level primitive extraction and organization of low level primitives through dynamic neural networks. These frameworks are compared and their relative advantages and disadvantages are outlined.
• M. Penas, M. J. Carreira, M. G. Penedo, C. Mariño, "Segment extraction using Burns principles in a pseudo-color fuzzy Hough transform", Lecture Notes in Computer Science: Pattern Recognition and Image Analysis, 3523(2), 175-182, 2005.
  [Abstract] [PDF] [+Info.]
  This paper describes a computational framework developed for the extraction of low-level directional primitives present in an image, and subsequent organization through a line segment detector. The system is divided in three stages: extraction of the directional features in the image through an efficient implementation of Gabor wavelet decomposition; reduction of these high dimensionality results by means of a growing cell structure; and extraction of the segments from the image. This last step was first implemented through a pseudo-color Fuzzy Hough Transform and then improved through some principles of the Burns segment detector.
• A. Pose-Reino, F. Gómez-Ulla, B. Hayik, M. J. Carreira, A. Mosquera, M. G. Penedo, "Computerized measuremente of retinal blood vessel calibre: description, validation and use to determine the influence og ageing and hypertension", Journal of hypertensión, 23(4), 843-50, 2005.
  [+Info.]

Year 2004

• I. G. Caderno, M. G. Penedo, C. Mariño, F. Gómez-Ulla, F. González, M. J. Carreira, "Automatic Extraction of the Retina AV Index ", Lecture Notes in Computer Science: Image Analysis and Recognition, 3212, 132-140, 2004.
  [+Info.]
• N. Barreira, M. G. Penedo, "Topological Active Volumes for Segmentation and Shape Reconstruction of Medical Images ", Lecture Notes in Computer Science: Image Analysis and Recognition, 3212, 43-50, 2004.
  [Abstract] [PDF] [+Info.]
  This paper presents a new methodology for automatic 3D segmentation and shape reconstruction of bones from tomographic cross-sections. This methodology uses the Topological Active Volumes model. The model is based on deformable models, it is able to integrate the most representative characteristics of the region-based and boundary-based segmentation models and it also provides information about the topological properties of the inside of detected objects. This model has the ability to perform topological local changes in its structure during the adjustment phase in order to: obtain a specific adjustment to object's local singularities, find several objects in the scene and identify and delimit holes in detected structures.
• C. Alonso-Montes, M. G. Penedo, "3D Object Surface Reconstruction Using Growing Self-organised Networks", Lecture Notes in Computer Science: Progress in Pattern Recognition, Image Analysis and Applications, 3287, 163-170, 2004.
  [Abstract] [PDF] [+Info.]
  This paper studies the adaptation of growing self-organised neural networks for 3D object surface reconstruction. Nowadays, input devices and filtering techniques obtain 3D point positions from the object surface without connectivity information. Growing self-organised networks can obtain the implicit surface mesh by means of a clustering process over the input data space maintaining at the same time the spatial-topology relations. The influence of using additional point features (e.g. gradient direction) as well as the methodology characterized in this paper have been studied to improve the obtained surface mesh.
• D. García-Pérez, A. Mosquera, M. Ortega, M. G. Penedo, "Significant Perceptual Regions by Active-Nets ", Lecture Notes in Computer Science: Image Analysis and Recognition, 3211, 795-802, 2004.
  [Abstract] [PDF] [+Info.]
  The available visual information is quickly growing now a days, it is the reason of the emerging of a new research field, oriented to the automatic retrieval of this kind of information. These systems usually uses perceptual features of the images (color, shape, texture,...). There is an important gap between the features used by the CBIR systems and the human perception of the information of an image. This work introduces a technique to extract significant perceptual regions of an image. The developed algorithm uses a bidimensional active model, active nets, these nets are guided by the chromatic components of a perceptual color space of the tested image. The restriction to only chromatic information made the fitting of an active net to the significant perceptual regions more tolerant to illumination problems of the image. The final objective will be to associate significant perceptual regions with semantic descriptors of the objects present in an image.
• M. Penas, M. J. Carreira, M. G. Penedo, M. Mirmehdi, B. T. Thomas, "Comparative Frameworks for Directional Primitive Extraction ", Lecture Notes in Computer Science: Image Analysis and Recognition, 3211, 212-219, 2004.
  [Abstract] [PDF] [+Info.]
  This paper introduces two alternative computational frameworks for the extraction of the directional primitives present in an image. Both frameworks are divided into three stages: low level primitive extractio organisation of low level primitives by means of dynamical neural netwo (growing cell structures) and segment extraction through a pseudo-colou Hough transform. The alternative frameworks are compared and their relative advantages and disadvantages are outlined.

Year 2003

• C. Mariño, M. G. Penedo, M. J. Carreira, F. González, "Retinal Angiography Based Authentication ", Lecture Notes in Computer Science: Progress in Pattern Recognition, Speech and Image Analysis, 2905, 306-313, 2003.
  [Abstract] [PDF] [+Info.]
  Traditional authentication (identity verification) systems, employed to gain access to a private area in a building or to data stored in a computer, are based on something the user has (an authentication card, a magnetic key) or something the user knows (a password, an identification code). But emerging technologies allow for more reliable and comfortable for the user, authentication methods, most of them based in biometric parameters. Much work could be found in literature about biometric based authentication, using parameters like iris, voice, fingerprint, face characteristics, and others. In this work a novel authentication method is presented, and first results obtained are shown. The biometric parameter employed for the authentication is the retinal vessel tree, acquired through a retinal angiography. It has already been asserted by expert clinicians that the configuration of the retinal vessels is unique for each individual and that it does not vary in his life, so it is a very well suited identification characteristic. Before the verification process can be executed, a registration step is needed to align both the reference image and the picture to be verified. A fast and reliable registration method is used to perform that step, so that the whole authentication process takes very little time.
• C. Mariño, M. G. Penedo, M. Penas, "Design of a distributed medical application using XML based data interchange ", Lecture Notes in Computer Science: Computer Aided Systems Theory, 2809, 253-263, 2003.
  [Abstract] [PDF] [+Info.]
  The recent proliferation of computing devices and the contexts in which they are used demand diversity in distributed applications as well. The objective of our research is the development of a medical framework where information from patients can be accessed from heterogeneous and (possibly) mobile computing environments. Moreover, high availability and reliability are also milestones in that system. The former objective is achieved by using eXtensible Markup Language (XML) for the communication medium, in combination with eXtensible Stylesheet Language (XSL) transformations to allow different kinds of clients access the data. High availability is achieved by using a concurrent and distributed language, Erlang/OTP, for the development on the server side. Also, in the server side, techniques coming from the formal methods area are applied to improve the system design and performance and to ensure the system correctness. And finally, reliability, confidentially and authentication, fundamental items in the data communications, are accomplished by mean of the Secure Socket Layer (SSL) protocol.
• M. Penas, M. G. Penedo, M. J. Carreira, "Gabor Wavelests and Auto-organised Structures for directional primitive extraction ", Lecture Notes in Computer Science: Pattern Recognition and Image Analysis, 2652, 722-732, 2003.
  [Abstract] [PDF] [+Info.]
  This paper describes a computational framework developed for the extraction of low-level directional primitives present in an image, and subsequent organization using the laws of perceptual grouping. The system is divided in two stages. The first one consists on the extraction of the direction of pixels in the image, through an efficient implementation of Gabor wavelet decomposition. The second one consists on the reduction of these high dimensionality results by means of an auto-organized structure. For this second stage, three different auto-organized structures have been studied: self-organized maps (SOM), growing cell structures (GCS) and growing neural gas (GNG). Results have showed that GCS is the most appropriate structure in the context of this work.
• M. Penas, M. J. Carreira, M. G. Penedo, "Perceptual organization of directional primitives using a pseudo-color Hough transform", Lecture Notes in Computer Science, 2749, 893-899, 2003.
  [Abstract] [PDF] [+Info.]
  This paper describes a computational framework developed for the extraction of low-level directional primitives present in an image, and subsequent organization using the laws of perceptual grouping. The system is divided in three stages. Extraction of the directional features in the image, through an efficient implementation of Gabor wavelet decomposition. Reduction of these high dimensionality results by means of growing cell structures. And extraction of the segments from the image by means of a Fuzzy Hough Transform.
• N. Barreira, M. G. Penedo, C. Mariño, F. M. Ansia, "Topological Active Volumes ", Lecture Notes in Computer Science: Computer Analysis of Images and Patterns, 2756, 337-344, 2003.
  [Abstract] [PDF] [+Info.]
  In the last years, deformable models raised much interest and found various applications in the field of 2D and 3D computer vision. Active surfaces are usually employed for segmentation and object reconstruction. In this paper, a new model for 3D image segmentation is proposed, the Topological Active Volumes (TAV). This model is based on deformable models, it is able to integrate the most representative characteristics of the region-based and boundary-based segmentation models and it also provides information about the topological properties of the inside of detected objects. This model has the ability to perform topological local changes in its structure during the adjustment phase in order to: obtain a specific adjustment to object's local singularities, find several objects in the scene and identify and delimit holes in detected structures.

Year 2002

• C. Mariño, M. Penas, M. G. Penedo, J. M. Barja, V. Leborán, M. J. Carreira, "Methodology for the Registration of Whole SLO sequences", Pattern Recognition; IEEE Computer Society Press, 1, 779-783, 2002.
  [Abstract] [PDF]
  Images acquired with the Scanning Laser Ophthalmoscope (SLO) device offer a very good tool for the evaluation of retinal diseases and obtaining interesting haemodynamic variables related to retinal blood flow. But because of the unavoidable eye movements, it is necessary a previous registration process to obtain a reliable analysis of the information extracted from these set of frames. This paper explains an algorithm which allows a fully automatic registration of the whole \SLO sequence by combining two registration techniques: a landmark based registration and a mutual information based one. The accurate registration of the frames will allow the automatic calculation of the so called dye dilution curves, which are computed by means of a robust fitting algorithm based on the optimisation method Simulated Annealing. With that curves it will be possible to obtain valuable haemodynamic variables, like the arteriovenous passage time described here.
• M. J. Carreira, M. Mirmehdi, B. T. Thomas, M. Penas, "Perceptual primitives from an extended 4D Hough transform", Image and Vision Computing, 20, 969-980, 2002.
  [Abstract] [PDF]
  Directional features extracted from Gabor wavelets responses were used to train a structure of self-organising maps, thus classifying each pixel in the image within a neuron-map. Resulting directional primitives were grouped into perceptual primitives introducing an extended 4D Hough transform to group pixels with similar directional features. These can then be used as perceptual primitives to detect salient structures. The proposed method has independently fixed parameters that do not need to be tuned for different kind or quality of images. We present results in application to noisy FLIR images and show that line primitives for complex structures, such as bridges, or simple structures, such as runways, can be found by this approach. We compare and demonstrate the quality of our results with those obtained through a parameter-dependent traditional Canny edge detector and Hough line finding process.

Year 2001

• M. Penas, M. J. Carreira, M. G. Penedo, "Autoorganized structures for extraction of directional primitives", Lecture Notes in Computer Science, 2085, 628-635, 2001.
  [Abstract] [PDF]
  In this work we have used directional features extracted from Gabor wavelet responses to compare different auto-organised networks in order to extract perceptual primitives without taking into account the kind of images to analyse. This is an adequate problem to prove the performance of these models because of the high dimensionality of the input space. Three different models have been analysed: self-organised maps, growing-cell structures and growing neural gas. Results have proved that growing-cell structures generalise better all directional perceptual primitives we are searching for, and they do not provide very noisy images.
• M. Penas, M. G. Penedo, M. J. Carreira, "Autoorganised Structures for Extraction of Perceptual Primitives", Lecture Notes in Computer Cience, Connectionist Models of Neurons, Learning Processes, and Artificial Intelligence , 628-636, 2001.
  [Abstract] [PDF] [+Info.]
  In this work we have used directional features extracted from Gabor wavelet responses to compare different auto-organized networks in order to extract perceptual primitives without taking into account the kind of images to analyse. This is an adequate problem to prove the performance of these models because of the high dimensionality of the input space. Three different models have been analysed: self-organized maps, growing cell structures and growing neural gas. Results have proved that growing-cell structures generalise better all directional perceptual primitives we are searching for, and they do not provide very noisy images.

Year 2000

• F. M. Ansia, M. G. Penedo, C. Mariño, A. Mosquera, J. López, "Automatic 3D Shape Reconstruction of Bones Using Active Nets Based Segmentation", 15th International Conference on Pattern Recognition, 1, 486-489, 2000.
  [PDF]

Year 1998

• M. G. Penedo, D. Cabello, A. Mosquera, M. J. Carreira, "Computed-Aided Diagnosis: A Neural Network Based Approach to Lung Nodule Detection ", IEEE Transactions on Medical Image , 17,6, 872-880, 1998.
  
• M. J. Carreira, D. Cabello, M. G. Penedo, A. Mosquera, "Computer Aided Diagnoses: Automatic Detection of Lung Nodules ", Medical Physics , 1998-2006, 1998.
  

Year 1997

• M. G. Penedo, "Diagnóstico Asistido por Ordenador. Aplicación de Redes Neuronales Artificiales en la Detección Automática de Nódulos Pulmonare s", Servicio de Publicaciones de la Universidad de Santiago de Compostela. Edición de Tesis de Doctorado (formato CD-ROM), 1997.
  

Year 1995

• D. Cabello, S. Barro, A. Mosquera, M. G. Penedo, J. M. Pardo, M. J. Carreira, "Aplicaciones de Redes Neuronales Artificiales en Procesado de Imágenes ", en Barro, S. y Mira J. (eds). Computación Neuronal, Servicio de Publicaciones e Intercambio Científico de la Universidad de Santiago de Compostela,, 457-477, 1995.
  
• M. G. Penedo, M. J. Carreira, D. Cabello, "Automatic Detection of Lung Nodules in Medical Images: Analysis in a Curvature Space ", en Lemke, H.U.; Inamura, K.; Jaffe, C.C. and Vannier, M.W. (eds). Computer Assisted Radiology, Springer, Berlín , 1276-1277, 1995.
  
• M. J. Carreira, D. Cabell