3D Maps Representation Using GNG

Current RGB-D sensors provide a big amount of valuable information for mobile robotics tasks like 3D map reconstruction, but the storage and processing of the incremental data provided by the different sensors through time quickly become unmanageable. In this work, we focus on 3D maps representation and propose the use of the Growing Neural Gas (GNG) network as a model to represent 3D input data. GNG method is able to represent the input data with a desired amount of neurons or resolution while preserving the topology of the input space. Experiments show how GNG method yields a better input space adaptation than other state-of-the-art 3D map representation methods.

Seguimiento del Máster en Ingeniería de Telecomunicación: medidas para conseguir la calidad y la excelencia

El Máster Universitario en Ingeniería de Telecomunicación en la Universidad de Alicante es el título que se imparte desde la Escuela Politécnica Superior y que habilita para el ejercicio de la profesión regulada de Ingeniero de Telecomunicación. Consta de 90 ECTS y se imparte a lo largo de 2 cursos académicos. El máster está implantado desde el curso 2011-2012 por lo que durante el actual curso 2012-2013 tendremos egresados de la primera promoción. Una vez implantado en su totalidad, es posible realizar un seguimiento del máster con la intención de obtener la excelencia académica mediante todas aquellas medidas y procesos de evaluación internos que sean necesarios. Además también es conveniente evaluar el impacto social, universitario y empresarial del máster evaluando las carencias, lagunas y oportunidades de formación que se detecten y tomando las oportunas medidas correctoras. Será necesario tener en cuenta la relación del título con el mundo empresarial y con la sociedad en general que tiene el sector de las telecomunicaciones en la evaluación de la calidad docente en el aula.

Valoración por los estudiantes del uso de foros Moodle en la evaluación docente

En la asignatura Sistemas Inteligentes del grado en Ingeniería Informática, durante el curso 2012-13, se planteó a los estudiantes un sistema de evaluación de la parte teórica consistente en la realización de controles on-line utilizando cuestionarios Moodle. Para la elaboración de los cuestionarios, se elegirían preguntas propuestas por los estudiantes en cada uno de los foros temáticos correspondientes. De cada tema se creó una base de preguntas Moodle para la generación con restricciones de cuestionarios aleatorios. El funcionamiento de los foros de aportación de preguntas partía de un foro creado para cada tema donde los estudiantes iban añadiendo un hilo para cada pregunta que proponían. En el hilo se producía el debate y participación de los estudiantes relativo al contenido y calidad de la pregunta. Una vez cerrado cada foro, los profesores seleccionaban las que cumplían requisitos de calidad, creando así la base de preguntas de los cuestionarios Moodle. La asignatura ha tenido más de 120 estudiantes, con 14 foros, lo que ha hecho que se superaran los 1.500 hilos de preguntas. En este artículo se presenta la encuesta de valoración por parte de los estudiantes, así como detalles de participación en los foros, los resultados obtenidos y las conclusiones. Deseamos destacar el soporte de la “Red de Investigación en Sistemas Inteligentes. Uso de los foros Moodle en la metodología docente” para realizar este estudio. Código de Red ICE: 2847.

Learning Probabilistic Features for Robotic Navigation Using Laser Sensors

SLAM is a popular task used by robots and autonomous vehicles to build a map of an unknown environment and, at the same time, to determine their location within the map. This paper describes a SLAM-based, probabilistic robotic system able to learn the essential features of different parts of its environment. Some previous SLAM implementations had computational complexities ranging from O(Nlog(N)) to O(N2), where N is the number of map features. Unlike these methods, our approach reduces the computational complexity to O(N) by using a model to fuse the information from the sensors after applying the Bayesian paradigm. Once the training process is completed, the robot identifies and locates those areas that potentially match the sections that have been previously learned. After the training, the robot navigates and extracts a three-dimensional map of the environment using a single laser sensor. Thus, it perceives different sections of its world. In addition, in order to make our system able to be used in a low-cost robot, low-complexity algorithms that can be easily implemented on embedded processors or microcontrollers are used.

Improving classification using a Confidence Matrix based on weak classifiers applied to OCR

This paper proposes a new feature representation method based on the construction of a Confidence Matrix (CM). This representation consists of posterior probability values provided by several weak classifiers, each one trained and used in different sets of features from the original sample. The CM allows the final classifier to abstract itself from discovering underlying groups of features. In this work the CM is applied to isolated character image recognition, for which several set of features can be extracted from each sample. Experimentation has shown that the use of CM permits a significant improvement in accuracy in most cases, while the others remain the same. The results were obtained after experimenting with four well-known corpora, using evolved meta-classifiers with the k-Nearest Neighbor rule as a weak classifier and by applying statistical significance tests.

Analyzing the commercial activities of a street network by ranking their nodes: a case study in Murcia, Spain

Urban researchers and planners are often interested in understanding how economic activities are distributed in urban regions, what forces influence their special pattern and how urban structure and functions are mutually dependent. In this paper, we want to show how an algorithm for ranking the nodes in a network can be used to understand and visualize certain commercial activities of a city. The first part of the method consists of collecting real information about different types of commercial activities at each location in the urban network of the city of Murcia, Spain. Four clearly differentiated commercial activities are studied, such as restaurants and bars, shops, banks and supermarkets or department stores, but obviously we can study other. The information collected is then quantified by means of a data matrix, which is used as the basis for the implementation of a PageRank algorithm which produces a ranking of all the nodes in the network, according to their significance within it. Finally, we visualize the resulting classification using a colour scale that helps us to represent the business network.

Improving kNN multi-label classification in Prototype Selection scenarios using class proposals

Prototype Selection (PS) algorithms allow a faster Nearest Neighbor classification by keeping only the most profitable prototypes of the training set. In turn, these schemes typically lower the performance accuracy. In this work a new strategy for multi-label classifications tasks is proposed to solve this accuracy drop without the need of using all the training set. For that, given a new instance, the PS algorithm is used as a fast recommender system which retrieves the most likely classes. Then, the actual classification is performed only considering the prototypes from the initial training set belonging to the suggested classes. Results show that this strategy provides a large set of trade-off solutions which fills the gap between PS-based classification efficiency and conventional kNN accuracy. Furthermore, this scheme is not only able to, at best, reach the performance of conventional kNN with barely a third of distances computed, but it does also outperform the latter in noisy scenarios, proving to be a much more robust approach.

Tuning compilations by multi-objective optimization: Application to Apache web server

Modern compilers present a great and ever increasing number of options which can modify the features and behavior of a compiled program. Many of these options are often wasted due to the required comprehensive knowledge about both the underlying architecture and the internal processes of the compiler. In this context, it is usual, not having a single design goal but a more complex set of objectives. In addition, the dependencies between different goals are difficult to be a priori inferred. This paper proposes a strategy for tuning the compilation of any given application. This is accomplished by using an automatic variation of the compilation options by means of multi-objective optimization and evolutionary computation commanded by the NSGA-II algorithm. This allows finding compilation options that simultaneously optimize different objectives. The advantages of our proposal are illustrated by means of a case study based on the well-known Apache web server. Our strategy has demonstrated an ability to find improvements up to 7.5% and up to 27% in context switches and L2 cache misses, respectively, and also discovers the most important bottlenecks involved in the application performance.

Using a RGB-D camera for 6DoF SLAM

This paper presents a method for fast calculation of the egomotion done by a robot using visual features. The method is part of a complete system for automatic map building and Simultaneous Localization and Mapping (SLAM). The method uses optical flow in order to determine if the robot has done a movement. If so, some visual features which do not accomplish several criteria (like intersection, unicity, etc,) are deleted, and then the egomotion is calculated. We use a state-of-the-art algorithm (TORO) in order to rectify the map and solve the SLAM problem. The proposed method provides better efficiency that other current methods.

An improvement of a SLAM RGB-D method with movement prediction derived from a study of visual features

This paper presents a method for the fast calculation of a robot’s egomotion using visual features. The method is part of a complete system for automatic map building and Simultaneous Location and Mapping (SLAM). The method uses optical flow to determine whether the robot has undergone a movement. If so, some visual features that do not satisfy several criteria are deleted, and then egomotion is calculated. Thus, the proposed method improves the efficiency of the whole process because not all the data is processed. We use a state-of-the-art algorithm (TORO) to rectify the map and solve the SLAM problem. Additionally, a study of different visual detectors and descriptors has been conducted to identify which of them are more suitable for the SLAM problem. Finally, a navigation method is described using the map obtained from the SLAM solution.

Deterministic mathematical morphology for CAD/CAM

Purpose – The purpose of this paper is to present a new geometric model based on the mathematical morphology paradigm, specialized to provide determinism to the classic morphological operations. The determinism is needed to model dynamic processes that require an order of application, as is the case for designing and manufacturing objects in CAD/CAM environments. Design/methodology/approach – The basic trajectory-based operation is the basis of the proposed morphological specialization. This operation allows the definition of morphological operators that obtain sequentially ordered sets of points from the boundary of the target objects, inexistent determinism in the classical morphological paradigm. From this basic operation, the complete set of morphological operators is redefined, incorporating the concept of boundary and determinism: trajectory-based erosion and dilation, and other morphological filtering operations. Findings – This new morphological framework allows the definition of complex three-dimensional objects, providing arithmetical support to generating machining trajectories, one of the most complex problems currently occurring in CAD/CAM. Originality/value – The model proposes the integration of the processes of design and manufacture, so that it avoids the problems of accuracy and integrity that present other classic geometric models that divide these processes in two phases. Furthermore, the morphological operative is based on points sets, so the geometric data structures and the operations are intrinsically simple and efficient. Another important value that no excessive computational resources are needed, because only the points in the boundary are processed.

A new boundary-based morphological model

Mathematical morphology addresses the problem of describing shapes in an n-dimensional space using the concepts of set theory. A series of standardized morphological operations are defined, and they are applied to the shapes to transform them using another shape called the structuring element. In an industrial environment, the process of manufacturing a piece is based on the manipulation of a primitive object via contact with a tool that transforms the object progressively to obtain the desired design. The analogy with the morphological operation of erosion is obvious. Nevertheless, few references about the relation between the morphological operations and the process of design and manufacturing can be found. The non-deterministic nature of classic mathematical morphology makes it very difficult to adapt their basic operations to the dynamics of concepts such as the ordered trajectory. A new geometric model is presented, inspired by the classic morphological paradigm, which can define objects and apply morphological operations that transform these objects. The model specializes in classic morphological operations, providing them with the determinism inherent in dynamic processes that require an order of application, as is the case for designing and manufacturing objects in professional computer-aided design and manufacturing (CAD/CAM) environments. The operators are boundary-based so that only the points in the frontier are handled. As a consequence, the process is more efficient and more suitable for use in CAD/CAM systems.

JavaVis: An Integrated Computer Vision Library for Teaching Computer Vision

In this article, we present a new framework oriented to teach Computer Vision related subjects called JavaVis. It is a computer vision library divided in three main areas: 2D package is featured for classical computer vision processing; 3D package, which includes a complete 3D geometric toolset, is used for 3D vision computing; Desktop package comprises a tool for graphic designing and testing of new algorithms. JavaVis is designed to be easy to use, both for launching and testing existing algorithms and for developing new ones.

Análisis y métodos numéricos con Geogebra

Cuaderno de prácticas de Matemáticas II, Grado en Ingeniería Informática, Universidad de Alicante.

Estudio y planificación de contenidos, materiales y metodologías docentes según el EEES:Itin. Gest. Contenidos. 4º Ingeniería Multimedia

La red pretende servir de plataforma para la coordinación de las asignaturas de cuarto curso que constituyen el itinerario de Gestión de Contenidos correspondiente al grado de Ingeniería Multimedia para la ejecución de una metodología basada en proyectos. La red ha detectado los problemas de coordinación asociados a gestionar simultáneamente proyectos de naturaleza variada y las posibilidades de conexión de estos mismos con las asignaturas de Prácticas de empresa I y II y la elaboración del Trabajo de Fin de Grado. Se proponen soluciones para paliar estos problemas y se explica el alcance de los objetivos a conseguir.