Artificial analysis of molecular marker loci linked to tree resistance response by an artificial neural network

One of the biggest challenges that software developers face is to make an accurate estimate of the project effort. Radial basis function neural networks have been used to software effort estimation in this work using NASA dataset. This paper evaluates and compares radial basis function versus a regression model. The results show that radial basis function neural network have obtained less Mean Square Error than the regression method.

Cognitive strategies for security in wireless sensor networks

Las redes de sensores inalámbricas son uno de los sectores con más crecimiento dentro de las redes inalámbricas. La rápida adopción de estas redes como solución para muchas nuevas aplicaciones ha llevado a un creciente tráfico en el espectro radioeléctrico. Debido a que las redes inalámbricas de sensores operan en las bandas libres Industrial, Scientific and Medical (ISM) se ha producido una saturación del espectro que en pocos años no permitirá un buen funcionamiento. Con el objetivo de solucionar este tipo de problemas ha aparecido el paradigma de Radio Cognitiva (CR). La introducción de las capacidades cognitivas en las redes inalámbricas de sensores permite utilizar estas redes para aplicaciones con unos requisitos más estrictos respecto a fiabilidad, cobertura o calidad de servicio. Estas redes que aúnan todas estas características son llamadas redes de sensores inalámbricas cognitivas (CWSNs). La mejora en prestaciones de las CWSNs permite su utilización en aplicaciones críticas donde antes no podían ser utilizadas como monitorización de estructuras, de servicios médicos, en entornos militares o de vigilancia. Sin embargo, estas aplicaciones también requieren de otras características que la radio cognitiva no nos ofrece directamente como, por ejemplo, la seguridad. La seguridad en CWSNs es un aspecto poco desarrollado al ser una característica no esencial para su funcionamiento, como pueden serlo el sensado del espectro o la colaboración. Sin embargo, su estudio y mejora es esencial de cara al crecimiento de las CWSNs. Por tanto, esta tesis tiene como objetivo implementar contramedidas usando las nuevas capacidades cognitivas, especialmente en la capa física, teniendo en cuenta las limitaciones con las que cuentan las WSNs. En el ciclo de trabajo de esta tesis se han desarrollado dos estrategias de seguridad contra ataques de especial importancia en redes cognitivas: el ataque de simulación de usuario primario (PUE) y el ataque contra la privacidad eavesdropping. Para mitigar el ataque PUE se ha desarrollado una contramedida basada en la detección de anomalías. Se han implementado dos algoritmos diferentes para detectar este ataque: el algoritmo de Cumulative Sum y el algoritmo de Data Clustering. Una vez comprobado su validez se han comparado entre sí y se han investigado los efectos que pueden afectar al funcionamiento de los mismos. Para combatir el ataque de eavesdropping se ha desarrollado una contramedida basada en la inyección de ruido artificial de manera que el atacante no distinga las señales con información del ruido sin verse afectada la comunicación que nos interesa. También se ha estudiado el impacto que tiene esta contramedida en los recursos de la red. Como resultado paralelo se ha desarrollado un marco de pruebas para CWSNs que consta de un simulador y de una red de nodos cognitivos reales. Estas herramientas han sido esenciales para la implementación y extracción de resultados de la tesis. ABSTRACT Wireless Sensor Networks (WSNs) are one of the fastest growing sectors in wireless networks. The fast introduction of these networks as a solution in many new applications has increased the traffic in the radio spectrum. Due to the operation of WSNs in the free industrial, scientific, and medical (ISM) bands, saturation has ocurred in these frequencies that will make the same operation methods impossible in the future. Cognitive radio (CR) has appeared as a solution for this problem. The networks that join all the mentioned features together are called cognitive wireless sensor networks (CWSNs). The adoption of cognitive features in WSNs allows the use of these networks in applications with higher reliability, coverage, or quality of service requirements. The improvement of the performance of CWSNs allows their use in critical applications where they could not be used before such as structural monitoring, medical care, military scenarios, or security monitoring systems. Nevertheless, these applications also need other features that cognitive radio does not add directly, such as security. The security in CWSNs has not yet been explored fully because it is not necessary field for the main performance of these networks. Instead, other fields like spectrum sensing or collaboration have been explored deeply. However, the study of security in CWSNs is essential for their growth. Therefore, the main objective of this thesis is to study the impact of some cognitive radio attacks in CWSNs and to implement countermeasures using new cognitive capabilities, especially in the physical layer and considering the limitations of WSNs. Inside the work cycle of this thesis, security strategies against two important kinds of attacks in cognitive networks have been developed. These attacks are the primary user emulator (PUE) attack and the eavesdropping attack. A countermeasure against the PUE attack based on anomaly detection has been developed. Two different algorithms have been implemented: the cumulative sum algorithm and the data clustering algorithm. After the verification of these solutions, they have been compared and the side effects that can disturb their performance have been analyzed. The developed approach against the eavesdropping attack is based on the generation of artificial noise to conceal information messages. The impact of this countermeasure on network resources has also been studied. As a parallel result, a new framework for CWSNs has been developed. This includes a simulator and a real network with cognitive nodes. This framework has been crucial for the implementation and extraction of the results presented in this thesis.

Fatiga en CFRP: Caracterización y Método Óptico para la Estimación del Daño

El desarrollo de nuevas estructuras aeroespaciales optimizadas, utilizan materiales compuestos, para los componentes críticos y subsistemas, principalmente polímeros reforzados con fibra de carbono (CFRP). Un conocimiento profundo del estado de daño por fatiga de estructuras de CFRP avanzado, es esencial para predecir la vida residual y optimizar los intervalos de inspección estructural, reparaciones y/o sustitución de componentes. Las técnicas actuales se basan principalmente en la medición de cargas estructurales a lo largo de la vida útil de la estructura mediante galgas extensométricas eléctricas. Con esos datos, se estima la vida a fatiga utilizando modelos de acumulación de daño. En la presente tesis, se evalúa la metodología convencional para la estimación de la vida a fatiga de un CFRP aeronáutico. Esta metodología está basada en la regla de acumulación de daño lineal de Palmgren-Miner, y es aplicada para determinar la vida a fatiga de estructuras sometidas a cargas de amplitud variable. Se ha realizado una campaña de ensayos con cargas de amplitud constante para caracterizar un CFRP aeronáutico a fatiga, obteniendo las curvas clásicas S-N, en diferentes relaciones de esfuerzo. Se determinaron los diagramas de vida constante, (CLD), también conocidos como diagramas de Goodman, utilizando redes neuronales artificiales debido a la ausencia de modelos coherentes para materiales compuestos. Se ha caracterizado la degradación de la rigidez debido al daño por fatiga. Se ha ensayado un segundo grupo de probetas con secuencias estandarizadas de cargas de amplitud variable, para obtener la vida a fatiga y la degradación de rigidez en condiciones realistas. Las cargas aplicadas son representativas de misiones de aviones de combate (Falstaff), y de aviones de transporte (Twist). La vida a fatiga de las probetas cicladas con cargas de amplitud variable, se comparó con el índice de daño teórico calculado en base a la regla de acumulación de daño lineal convencional. Los resultados obtenidos muestran predicciones no conservativas. Esta tesis también presenta el estudio y desarrollo, de una nueva técnica de no contacto para evaluar el estado de daño por fatiga de estructuras de CFRP por medio de cambios de los parámetros de rugosidad. La rugosidad superficial se puede medir fácilmente en campo con métodos sin contacto, mediante técnicas ópticas tales como speckle y perfilómetros ópticos. En el presente estudio, se han medido parámetros de rugosidad superficial, y el factor de irregularidad de la superficie, a lo largo de la vida de las probetas cicladas con cargas de amplitud constante y variable, Se ha obtenido una buena tendencia de ajuste al correlacionar la magnitud de la rugosidad y el factor de irregularidad de la superficie con la degradación de la rigidez de las probetas fatigadas. Estos resultados sugieren que los cambios en la rugosidad superficial medida en zonas estratégicas de componentes y estructuras hechas de CFRP, podrían ser indicativas del nivel de daño interno debido a cargas de fatiga. Los resultados también sugieren que el método es independiente del tipo de carga de fatiga que ha causado el daño. Esto último hace que esta técnica de medición sea aplicable como inspección para una amplia gama de estructuras de materiales compuestos, desde tanques presurizados con cargas de amplitud constante, estructuras aeronáuticas como alas y colas de aeronaves cicladas con cargas de amplitud variable, hasta aplicaciones industriales como automoción, entre otros. ABSTRACT New optimized aerospace structures use composite materials, mainly carbon fiber reinforced polymer composite (CFRP), for critical components and subsystems. A strong knowledge of the fatigue state of highly advanced (CFRP) structures is essential to predict the residual life and optimize intervals of structural inspection, repairs, and/or replacements. Current techniques are based mostly on measurement of structural loads throughout the service life by electric strain gauge sensors. These sensors are affected by extreme environmental conditions and by fatigue loads in such a way that the sensors and their systems require exhaustive maintenance throughout system life. In the present thesis, the conventional methodology based on linear damage accumulation rules, applied to determine the fatigue life of structures subjected to variable amplitude loads was evaluated for an aeronautical CFRP. A test program with constant amplitude loads has been performed to obtain the classical S-N curves at different stress ratios. Constant life diagrams, CLDs, where determined by means of Artificial Neural Networks due to the absence of consistent models for composites. The stiffness degradation due to fatigue damage has been characterized for coupons under cyclic tensile loads. A second group of coupons have been tested until failure with a standardized sequence of variable amplitude loads, representative of missions for combat aircraft (Falstaff), and representative of commercial flights (Twist), to obtain the fatigue life and the stiffness degradation under realistic conditions. The fatigue life of the coupons cycled with variable amplitude loads were compared to the theoretical damage index calculated based on the conventional linear damage accumulation rule. The obtained results show non-conservative predictions. This thesis also presents the evaluation of a new non-contact technique to evaluate the fatigue damage state of CFRP structures by means of measuring roughness parameters to evaluate changes in the surface topography. Surface roughness can be measured easily on field with non-contact methods by optical techniques such as speckle and optical perfilometers. In the present study, surface roughness parameters, and the surface irregularity factor, have been measured along the life of the coupons cycled with constant and variable amplitude loads of different magnitude. A good agreement has been obtained when correlating the magnitude of the roughness and the surface irregularity factor with the stiffness degradation. These results suggest that the changes on the surface roughness measured in strategic zones of components and structures made of CFRP, could be indicative of the level of internal damage due to fatigue loads. The results also suggest that the method is independent of the type of fatigue load that have caused the damage. It makes this measurement technique applicable for a wide range of inspections of composite materials structures, from pressurized tanks with constant amplitude loads, to variable amplitude loaded aeronautical structures like wings and empennages, up to automotive and other industrial applications.

Anteroposterior neural tissue specification by activin-induced mesoderm

The transforming growth factor β superfamily member, activin, is able to induce mesodermal tissues in animal cap explants from Xenopus laevis blastula stage embryos. Activin can act like a morphogen of the dorsoventral axis in that lower doses induce more ventral, and higher doses more dorsal, tissue types. Activin has also previously been reported to induce neural tissues in animal caps. From cell mixing experiments it was inferred that this might be an indirect effect of induced mesoderm signaling to uninduced ectoderm. Here we demonstrate directly that neural tissues do indeed arise by the action of induced mesoderm on uninduced ectoderm. Dorsal mesoderm is itself subdivided into posterior and anterior domains in vivo, but this had not been demonstrated for induced mesoderm. We therefore tested whether different concentrations of activin recreate these different anteroposterior properties as well. We show that the anteroposterior positional value of induced mesoderm, including its neuroinductive properties, depends on the dose of activin applied to the mesoderm, with lower doses inducing more posterior and higher doses giving more anterior markers. We discuss the implications of these results for patterning signals and the relationship between anteroposterior and dorsoventral axes.

Neuregulin-3 (NRG3): A novel neural tissue-enriched protein that binds and activates ErbB4

We describe the identification of Neuregulin-3 (NRG3), a novel protein that is structurally related to the neuregulins (NRG1). The NRG1/neuregulins are a diverse family of proteins that arise by alternative splicing from a single gene. These proteins play an important role in controlling the growth and differentiation of glial, epithelial, and muscle cells. The biological effects of NRG1 are mediated by receptor tyrosine kinases ErbB2, ErbB3, and ErbB4. However, genetic studies have suggested that the activity of ErbB4 may also be regulated in the central nervous system by a ligand distinct from NRG1. NRG3 is predicted to contain an extracellular domain with an epidermal growth factor (EGF) motif, a transmembrane domain, and a large cytoplasmic domain. We show that the EGF-like domain of NRG3 binds to the extracellular domain of ErbB4 in vitro. Moreover, NRG3 binds to ErbB4 expressed on cells and stimulates tyrosine phosphorylation of this receptor. The expression of NRG3 is highly restricted to the developing and adult nervous system. These data suggest that NRG3 is a novel, neural-enriched ligand for ErbB4.

Deciphering a neural code for vision

Deciphering the information that eyes, ears, and other sensory organs transmit to the brain is important for understanding the neural basis of behavior. Recordings from single sensory nerve cells have yielded useful insights, but single neurons generally do not mediate behavior; networks of neurons do. Monitoring the activity of all cells in a neural network of a behaving animal, however, is not yet possible. Taking an alternative approach, we used a realistic cell-based model to compute the ensemble of neural activity generated by one sensory organ, the lateral eye of the horseshoe crab, Limulus polyphemus. We studied how the neural network of this eye encodes natural scenes by presenting to the model movies recorded with a video camera mounted above the eye of an animal that was exploring its underwater habitat. Model predictions were confirmed by simultaneously recording responses from single optic nerve fibers of the same animal. We report here that the eye transmits to the brain robust “neural images” of objects having the size, contrast, and motion of potential mates. The neural code for such objects is not found in ambiguous messages of individual optic nerve fibers but rather in patterns of coherent activity that extend over small ensembles of nerve fibers and are bound together by stimulus motion. Integrative properties of neurons in the first synaptic layer of the brain appear well suited to detecting the patterns of coherent activity. Neural coding by this relatively simple eye helps explain how horseshoe crabs find mates and may lead to a better understanding of how more complex sensory organs process information.

A model of multiplicative neural responses in parietal cortex.

Visual responses of neurons in parietal area 7a are modulated by a combined eye and head position signal in a multiplicative manner. Neurons with multiplicative responses can act as powerful computational elements in neural networks. In the case of parietal cortex, multiplicative gain modulation appears to play a crucial role in the transformation of object locations from retinal to body-centered coordinates. It has proven difficult to uncover single-neuron mechanisms that account for neuronal multiplication. Here we show that multiplicative responses can arise in a network model through population effects. Specifically, neurons in a recurrently connected network with excitatory connections between similarly tuned neurons and inhibitory connections between differently tuned neurons can perform a product operation on additive synaptic inputs. The results suggest that parietal responses may be based on this architecture.

Endothelin-B receptor is expressed by neural crest cells in the avian embryo.

Disruptions of the genes encoding endothelin 3 (EDN3) and its receptor endothelin-B receptor (EDNRB) in the mouse result in defects of two neural crest (NC)-derived lineages, the melanocytes, and the enteric nervous system. To assess the mechanisms through which the EDN3/EDNRB signaling pathway can selectively act on these NC derivatives, we have studied the spatiotemporal expression pattern of the EDNRB gene in the avian embryo, a model in which NC development has been extensively studied. For this purpose, we have cloned the quail homologue of the mammalian EDNRB cDNA. EDNRB transcripts are present in NC cells before and during their emigration from the neural tube at all levels of the neuraxis. At later developmental stages, the receptor remains abundantly expressed in the peripheral nervous system including the enteric nervous system. In a previous study, we have shown that EDN3 enhances dramatically the proliferation of NC cells when they are at the pluripotent stage. We propose that the selective effect of EDN3 or EDNRB gene inactivation is due to the fact that both melanocytes and enteric nervous system precursors have to colonize large embryonic areas (skin and bowel) from a relatively small population of precursors that have to expand considerably in number. It is therefore understandable that a deficit in one of the growth-promoting pathways of NC cells has more deleterious effects on long-range migrating cells than on the NC derivatives which develop close to the neural primordium like the sensory and sympathetic ganglia.

Neuregulins with an Ig-like domain are essential for mouse myocardial and neuronal development.

Neuregulins are ligands for the erbB family of receptor tyrosine kinases and mediate growth and differentiation of neural crest, muscle, breast cancer, and Schwann cells. Neuregulins contain an epidermal growth factor-like domain located C-terminally to either an Ig-like domain or a cysteine-rich domain specific to the sensory and motor neuron-derived isoform. Here it is shown that elimination of the Ig-like domain-containing neuregulins by homologous recombination results in embryonic lethality associated with a deficiency of ventricular myocardial trabeculation and impairment of cranial ganglion development. The erbB receptors are expressed in myocardial cells and presumably mediate the neuregulin signal originating from endocardial cells. The trigeminal ganglion is reduced in size and lacks projections toward the brain stem and mandible. We conclude that IgL-domain-containing neuregulins play a major role in cardiac and neuronal development.

Homophilic adhesion mediated by the neural cell adhesion molecule involves multiple immunoglobulin domains.

The neural cell adhesion molecule (N-CAM) mediates homophilic binding between a variety of cell types including neurons, neurons and glia, and neurons and muscle cells. The mechanism by which N-CAM on one cell interacts with N-CAM on another, however, is unknown. Attempts to identify which of the five immunoglobulin-like domains (Ig I-V) and the two fibronectin type III repeats (FnIII 1-2) in the extracellular region of N-CAM are involved in this process have led to ambiguous results. We have generated soluble recombinant proteins corresponding to each of the individual immunoglobulin domains and the combined FnIII 1-2 and prepared polyclonal antibodies specific for each. The purified proteins and antibodies were used in aggregation experiments with fluorescent microspheres and chicken embryo brain cells to determine possible contributions of each domain to homophilic adhesion. The recombinant domains were tested for their ability to bind to purified native N-CAM, to bind to each other, and to inhibit the aggregation of N-CAM on microspheres and the aggregation of neuronal cells. Each of the immunoglobulin domains bound to N-CAM, and in solution all of the immunoglobulin domains inhibited the aggregation of N-CAM-coated microspheres. Soluble Ig II, Ig III, and Ig IV inhibited neuronal aggregation; antibodies against whole N-CAM, the Ig III domain, and the Ig I domain all strongly inhibited neuronal aggregation, as well as the aggregation of N-CAM-coated microspheres. Of all the domains, the third immunoglobulin domain alone demonstrated the ability to self-aggregate, whereas Ig I bound to Ig V and Ig II bound to Ig IV. The combined FnIII 1-2 exhibited a slight ability to self-aggregate but did not bind to any of the immunoglobulin-like domains. These results suggest that N-CAM-N-CAM binding involves all five immunoglobulin domains and prompt the hypothesis that in homophilic cell-cell binding mediated by N-CAM these domains may interact pairwise in an antiparallel orientation.

Neural cell adhesion molecule (N-CAM) inhibits astrocyte proliferation after injury to different regions of the adult rat brain.

After a penetrating lesion in the central nervous system, astrocytes enlarge, divide, and participate in creating an environment that adversely affects neuronal regeneration. We have recently shown that the neural cell adhesion molecule (N-CAM) partially inhibits the division of early postnatal rat astrocytes in vitro. In the present study, we demonstrate that addition of N-CAM, the third immunoglobulin-like domain of N-CAM, or a synthetic decapeptide corresponding to a putative homophilic binding site in N-CAM partially inhibits astrocyte proliferation after a stab lesion in the adult rat brain. Animals were lesioned in the cerebral cortex, hippocampus, or striatum with a Hamilton syringe and needle at defined stereotaxic positions. On one side, the lesions were concomitantly infused with N-CAM or with one of the N-CAM-related molecules. As a control, a peptide of the same composition as the N-CAM decapeptide but of random sequence was infused on the contralateral side of the brain. We consistently found that the population of dividing astrocytes was significantly smaller on the side in which N-CAM or one of the N-CAM-related molecules was infused than on the opposite side. The inhibition was greatest in the cortical lesion sites (approximately 50%) and was less pronounced in the hippocampus (approximately 25%) and striatum (approximately 20%). Two weeks after the lesion, the cerebral cortical sites infused with N-CAM continued to exhibit a significantly smaller population of dividing astrocytes than the sites on the opposite side. When N-CAM and basic fibroblast growth factor, which is known to stimulate astrocyte division in vitro, were coinfused into cortical lesion sites, astrocyte proliferation was still inhibited. These results suggest the hypothesis that, by reducing glial proliferation, N-CAM or its peptides may help create an environment that is more suitable for neuronal regeneration.

A conserved family of elav-like genes in vertebrates.

A large family of genes encodes proteins with RNA recognition motifs that are presumed to bind RNA and to function in posttranscriptional regulation. Neural-specific members of this family include elav, a gene required for correct differentiation and maintenance of neurons in Drosophila melanogaster, and a related gene, HuD, which is expressed in human neuronal cells. I have identified genes related to elav and HuD in Xenopus laevis, zebrafish, and mouse that define a family of four closely related vertebrate elav-like genes (elrA, elrB, elrC, and elrD) in fish, frogs, and mammals. In addition to protein sequence conservation, a segment of the 3'-untranslated sequence of elrD is also conserved, implying a functional role in elrD expression. In adult frogs, elrC and elrD are exclusively expressed in the brain, whereas elrB is expressed in brain, testis, and ovary. During Xenopus development, elrC and elrD RNAs are detected by late gastrula and late neurula stages, respectively, whereas a nervous system-specific elrB RNA species is expressed by early tadpole stage. Additional elrB transcripts are detected in the ovary and early embryo, demonstrating a maternal supply of mRNA and possibly of protein. These expression patterns suggest a role for different elav-like genes in early development and neuronal differentiation. Surprisingly, elrA is expressed in all adult tissues tested and at all times during development. Thus, the widely expressed elrA is expected to have a related function in all cells.

Aplicação de uma rede neural artificial para a avaliação da rugosidade e soprosidade vocal

A avaliação perceptivo-auditiva tem papel fundamental no estudo e na avaliação da voz, no entanto, por ser subjetiva está sujeita a imprecisões e variações. Por outro lado, a análise acústica permite a reprodutibilidade de resultados, porém precisa ser aprimorada, pois não analisa com precisão vozes com disfonias mais intensas e com ondas caóticas. Assim, elaborar medidas que proporcionem conhecimentos confiáveis em relação à função vocal resulta de uma necessidade antiga dentro desta linha de pesquisa e atuação clínica. Neste contexto, o uso da inteligência artificial, como as redes neurais artificiais, indica ser uma abordagem promissora. Objetivo: Validar um sistema automático utilizando redes neurais artificiais para a avaliação de vozes rugosas e soprosas. Materiais e métodos: Foram selecionadas 150 vozes, desde neutras até com presença em grau intenso de rugosidade e/ou soprosidade, do banco de dados da Clínica de Fonoaudiologia da Faculdade de Odontologia de Bauru (FOB/USP). Dessas vozes, 23 foram excluídas por não responderem aos critérios de inclusão na amostra, assim utilizaram-se 123 vozes. Procedimentos: avaliação perceptivo-auditiva pela escala visual analógica de 100 mm e pela escala numérica de quatro pontos; extração de características do sinal de voz por meio da Transformada Wavelet Packet e dos parâmetros acústicos: jitter, shimmer, amplitude da derivada e amplitude do pitch; e validação do classificador por meio da parametrização, treino, teste e avaliação das redes neurais artificiais. Resultados: Na avaliação perceptivo-auditiva encontrou-se, por meio do teste Coeficiente de Correlação Intraclasse (CCI), concordâncias inter e intrajuiz excelentes, com p = 0,85 na concordância interjuízes e p variando de 0,87 a 0,93 nas concordâncias intrajuiz. Em relação ao desempenho da rede neural artificial, na discriminação da soprosidade e da rugosidade e dos seus respectivos graus, encontrou-se o melhor desempenho para a soprosidade no subconjunto composto pelo jitter, amplitude do pitch e frequência fundamental, no qual obteve-se taxa de acerto de 74%, concordância excelente com a avaliação perceptivo-auditiva da escala visual analógica (0,80 no CCI) e erro médio de 9 mm. Para a rugosidade, o melhor subconjunto foi composto pela Transformada Wavelet Packet com 1 nível de decomposição, jitter, shimmer, amplitude do pitch e frequência fundamental, no qual obteve-se 73% de acerto, concordância excelente (0,84 no CCI), e erro médio de 10 mm. Conclusão: O uso da inteligência artificial baseado em redes neurais artificiais na identificação, e graduação da rugosidade e da soprosidade, apresentou confiabilidade excelente (CCI > 0,80), com resultados semelhantes a concordância interjuízes. Dessa forma, a rede neural artificial revela-se como uma metodologia promissora de avaliação vocal, tendo sua maior vantagem a objetividade na avaliação.

A neural network algorithm to simplify 2D meshes

In this paper we propose a neural network model to simplify and 2D meshes. This model is based on the Growing Neural Gas model and is able to simplify any mesh with different topologies and sizes. A triangulation process is included with the objective to reconstruct the mesh. This model is applied to some problems related to urban networks.

Using GNG to improve 3D feature extraction—Application to 6DoF egomotion

Several recent works deal with 3D data in mobile robotic problems, e.g. mapping or egomotion. Data comes from any kind of sensor such as stereo vision systems, time of flight cameras or 3D lasers, providing a huge amount of unorganized 3D data. In this paper, we describe an efficient method to build complete 3D models from a Growing Neural Gas (GNG). The GNG is applied to the 3D raw data and it reduces both the subjacent error and the number of points, keeping the topology of the 3D data. The GNG output is then used in a 3D feature extraction method. We have performed a deep study in which we quantitatively show that the use of GNG improves the 3D feature extraction method. We also show that our method can be applied to any kind of 3D data. The 3D features obtained are used as input in an Iterative Closest Point (ICP)-like method to compute the 6DoF movement performed by a mobile robot. A comparison with standard ICP is performed, showing that the use of GNG improves the results. Final results of 3D mapping from the egomotion calculated are also shown.