Processing of Byzantine Neume Notation in Ancient Historical Manuscripts

This article presents the principal results of the doctoral thesis “Recognition of neume notation in historical documents” by Lasko Laskov (Institute of Mathematics and Informatics at Bulgarian Academy of Sciences), successfully defended before the Specialized Academic Council for Informatics and Mathematical Modelling on 07 June 2010.

Data Extraction from Carte Du Ciel Triple Images

Carte du Ciel (from French, map of the sky) is a part of a 19th century extensive international astronomical project whose goal was to map the entire visible sky. The results of this vast effort were collected in the form of astrographic plates and their paper representatives that are called astrographic maps and are widely distributed among many observatories and astronomical institutes over the world. Our goal is to design methods and algorithms to automatically extract data from digitized Carte du Ciel astrographic maps. This paper examines the image processing and pattern recognition techniques that can be adopted for automatic extraction of astronomical data from stars’ triple expositions that can aid variable stars detection in Carte du Ciel maps.

The impact of ageing and gender on visual mental imagery processes:a study of performance on tasks from the Complete Visual Mental Imagery Battery (CVMIB)

In this study we aim to evaluate the impact of ageing and gender on different visual mental imagery processes. Two hundred and fifty-one participants (130 women and 121 men; age range = 18–77 years) were given an extensive neuropsychological battery including tasks probing the generation, maintenance, inspection, and transformation of visual mental images (Complete Visual Mental Imagery Battery, CVMIB). Our results show that all mental imagery processes with the exception of the maintenance are affected by ageing, suggesting that other deficits, such as working memory deficits, could account for this effect. However, the analysis of the transformation process, investigated in terms of mental rotation and mental folding skills, shows a steeper decline in mental rotation, suggesting that age could affect rigid transformations of objects and spare non-rigid transformations. Our study also adds to previous ones in showing gender differences favoring men across the lifespan in the transformation process, and, interestingly, it shows a steeper decline in men than in women in inspecting mental images, which could partially account for the mixed results about the effect of ageing on this specific process. We also discuss the possibility to introduce the CVMIB in clinical assessment in the context of theoretical models of mental imagery.

Feasibility of MR-Based Body Composition Analysis in Large Scale Population Studies

Introduction Quantitative and accurate measurements of fat and muscle in the body are important for prevention and diagnosis of diseases related to obesity and muscle degeneration. Manually segmenting muscle and fat compartments in MR body-images is laborious and time-consuming, hindering implementation in large cohorts. In the present study, the feasibility and success-rate of a Dixon-based MR scan followed by an intensity-normalised, non-rigid, multi-atlas based segmentation was investigated in a cohort of 3,000 subjects. Materials and Methods 3,000 participants in the in-depth phenotyping arm of the UK Biobank imaging study underwent a comprehensive MR examination. All subjects were scanned using a 1.5 T MR-scanner with the dual-echo Dixon Vibe protocol, covering neck to knees. Subjects were scanned with six slabs in supine position, without localizer. Automated body composition analysis was performed using the AMRA Profiler™ system, to segment and quantify visceral adipose tissue (VAT), abdominal subcutaneous adipose tissue (ASAT) and thigh muscles. Technical quality assurance was performed and a standard set of acceptance/rejection criteria was established. Descriptive statistics were calculated for all volume measurements and quality assurance metrics. Results Of the 3,000 subjects, 2,995 (99.83%) were analysable for body fat, 2,828 (94.27%) were analysable when body fat and one thigh was included, and 2,775 (92.50%) were fully analysable for body fat and both thigh muscles. Reasons for not being able to analyse datasets were mainly due to missing slabs in the acquisition, or patient positioned so that large parts of the volume was outside of the field-of-view. Discussion and Conclusions In conclusion, this study showed that the rapid UK Biobank MR-protocol was well tolerated by most subjects and sufficiently robust to achieve very high success-rate for body composition analysis. This research has been conducted using the UK Biobank Resource.

Rôle de la conformation des glycoprotéines de l’enveloppe du VIH-1 dans la réponse cytotoxique cellulaire dépendante des anticorps et impact des protéines virales Nef et Vpu

Alors que d’énormes efforts sont mis de l’avant pour mettre en place des stratégies thérapeutiques contre l’infection au VIH-1, il est nécessaire de mieux cerner les déterminants viraux qui aideront à l’efficacité de celles-ci. En ce sens, une volumineuse littérature scientifique suggère que les anticorps contre le VIH-1 possédant une capacité à induire une réponse effectrice dépendante de leur portion Fc puissent jouer un rôle important dans la prévention de l’infection et dans la progression de la maladie. Cependant, peu d’information est disponible concernant les déterminants reconnus par ces anticorps et comment le virus s’en protège. Le but des travaux présentés dans cette thèse est donc d’élucider les mécanismes viraux contrôlant la reconnaissance des cellules infectées par ces anticorps capables d’induire une réponse effectrice. De par les corrélats de protection identifiés au cours de l’essai vaccinal RV144, les travaux présentés ici se concentrent sur la réponse cytotoxique dépendante des anticorps (ADCC), puisqu’il s’agit d’une réponse effectrice suggérée pour avoir joué un rôle dans la protection observée dans le RV144, seul essai vaccinal anti-VIH à avoir démontré un certain degré de protection. De plus, plusieurs anticorps capables d’induire cette réponse contre le VIH sont connus pour reconnaître les glycoprotéines de surface du virus (Env) dans une conformation dite ouverte, c’est-à-dire la conformation adoptée lors de la liaison d’Env avec son récepteur CD4 (épitopes CD4i). Nous avons mis au point deux techniques in vitro permettant d’étudier ces changements de conformation ainsi que leur impact sur la réponse ADCC. Les techniques mises au point, un ÉLISA sur base cellulaire pour mesurer les changements de conformation d’Env ainsi que la mesure de la réponse ADCC par cytométrie en flux, nous ont permis de démontrer comment le virus empêche l’exposition des épitopes d’Env CD4i. L’activité simultanée des protéines accessoires virales Nef et Vpu sur le retrait du récepteur CD4 de la surface des cellules infectées et l’inhibition du facteur de restriction Tétherine / BST-2 par Vpu contrôlent à la fois les niveaux d’Env et de CD4 à la surface cellulaire et donc modulent l’interaction Env-CD4 et ultimement la susceptibilité à la réponse ADCC contre les épitopes CD4i reconnus par des anticorps hautement prévalents lors de l’infection au VIH. Également, nous démontrons comment de petits composés mimant la liaison de CD4 sur Env sont capables de forcer l’exposition des épitopes CD4i, même en présence des protéines Nef et Vpu, et donc d’augmenter la susceptibilité des cellules infectées à la réponse ADCC. Une autre découverte présentée ici est la démonstration que la portion soluble d’Env produite par les cellules infectées peut interagir avec le récepteur CD4 des cellules non-infectées avoisinantes et induire leur reconnaissance et élimination par la réponse ADCC contre Env. Somme toute, la modulation de la réponse ADCC par l’interaction Env–CD4 représente un important pilier de la relation hôte – pathogène du VIH-1 de la perspective des réponses Fc-dépendantes. Les travaux présentés dans cette thèse ont le potentiel d’être utilisés dans l’élaboration de nouvelles stratégies antivirales tout en élargissant les connaissances fondamentales de cette interaction hôte – pathogène.

Riffe und fazielle Entwicklung der florigemma-Bank (Korallenoolith, Oxfordium) im Süntel und östlichen Wesergebirge (NW-Deutschland

During the Sedimentation of the platform carbonate deposits of the Korallenoolith Formation (middle Oxfordian to early Kimmeridgian) small buildups ofcorals formed in the Lower Saxony Basin. These bioconstructions are restricted to particular horizons (Untere Korallenbank,ßorigenuna-Bank Member etc.) and represent patch reefs and biostromes. In this study, the development of facies, fossil assemblages, spatial distribution of fossils, and reefs of the ßorigenuna-Bank Member (upper Middle Oxfordian) in the Süntel Mts and the eastern Wesergebirge Mts is described; the formation of reefs is discussed in detail. Twelve facies types are described and interpreted. They vary between high-energy deposits as well winnowed oolites and quiet-water lagoonal mudstones. Owing to the significance of biota, micro- and macrofossils are systematically described. The reefs are preserved in growth position, are characterized by numerous corresponding features and belong to a certain reef type. According to their size, shape and framework, they represent patch reefs, coral knobs (sensu James, 1983), coral thrombolite reefs (sensu Leinfelder et al., 1994) or “Klein- and Mitteldickichte” (sensu Laternser, 2001). Their growth fabric corresponds to the superstratal (dense) pillarstone (sensu Insalaco, 1998). As the top of the ßorigenuna-Bank displays an erosional unconformity (so-called Hauptdiskontinuität), the top of the reefs are erosionally capped. Their maximum height amounts to at least the maximum thickness of the ßorigenuna-Bank which does not exceed 4 metres. The diversity of coral fauna of the reefs is relatively low; a total of 13 species is recorded. The coral community is over- whelmingly dominated by the thin-branched ramose Thamnasteria dendroidea (Lamouroux) that forms aggregations of colonies (77?. dendroidea thickets). Leafy to platy Fungiastrea arachnoides (Parkinson) and Thamnasteria concinna (Goldfuss) occur subordinately, other species are only of minor importance. In a few cases, the reef-core consisting of Th. dendroidea thickets is laterally encrusted by platy F. arachnoides and Th. concinna colonies, and microbial carbonates. This zonation reflects probably a succession of different reef builders as a result of changing environmental conditions (allogenic succession). Moreover, some reefs are overlain by a biostrome made of large Solenopora jurassica nodules passing laterally in a nerinean bed. Mikrobial carbonates promoted reef growth and favoured the preservation of reef organismn in their growth position or in situ. They exhibit a platy, dendroid, or reticulate growth form or occur as downward-facing hemispheroids. According to their microstructure, they consist of a peloidal, clotted, or unstructured fabric (predominately layered and poorly structured thrombolite as well as clotted leiolite) (sensu Schmid, 1996). Abundant endo- and epibiontic organisms (bivalves, gastropods, echinoids, asteroids, ophiuroids, crabs etc) are linked to the reefs. With regard to their guild structure, the reefs represent occurrences at which only a few coral species serve as builder. Moreover, microbial carbonates contribute to both building and binding of the reefs. Additional binder as well as baffler are present, but not abundant. According to the species diversity, the dweller guild comprises by far the highest number of invertebrate taxa. The destroyer guild chiefly encompasses bivalves. The composition of the reef community was influenced by the habitat structure of the Th. dendroidea thickets. Owing to the increase in encrusting organisms and other inhabitants of the thickets, the locational factors changed, since light intensity and hydrodynamic energy level and combined parameters as oxygen supply declined in the crowded habitat. Therefore a characteristic succession of organisms is developed that depends on and responds to changing environmental conditions („community replacement sequence“). The succession allows the differentiation of different stages. It started after the cessation of the polyps with boring organisms and photoautotrophic micro-encrusters (calcareous algae, Lithocodium aggregatum). Following the death of these pioneer organisms, encrusting and adherent organisms (serpulids, „Terebella“ species, bryozoans, foraminifers, thecideidinids, sklerospongid and pharetronid sponges, terebratulids), small mobile organisms (limpets), and microbial induced carbonates developed. The final stage in the community replacement sequence gave rise to small cryptic habitats and organisms that belong to these caves (cryptobionts, coelobites). The habitat conditions especially favoured small non-rigid demosponges (“soft sponges”) that tolerate reduced water circulation. Reef rubble is negligible, so that the reefs are bordered by fossiliferous micritic limestone passing laterally in micritic limestone. Approximately 10% of the study area (outcropping florigemma-Bank) corresponds to reefal deposits whereas the remaining 90% encompass lagoonal inter-reefal deposits. The reef development is a good example for the interaction between reef growth, facies development and sea-level changes. It was initiated by a sea-level rise (transgression) and corresponding decrease in the hydrodynamic energy level. Colonization and reef growth took place on a coarse-grained Substrate composed of oncoids, larger foraminifers and bioclasts. Reef growth took place in a calm marine lagoonal setting. Increasing abundance of spherical coral morphs towards the Northeast (section Kessiehausen, northwestem Süntel Mts) reflects higher turbidity and a facies transition to coral occurrences of the ßorigenuna-Bank Member in the adjacent Deister Mts. The reef growth was neither influenced by stonns nor by input of siliciclastic deposits, and took place in short time - probably in only a thousand years under most probably mesotrophic conditions. The mass appearance of solenoporids and nerineids in the upper part of the ßorigenuna-Bank Member point to enhanced nutrient level as a result of regression. In addition, this scenario of fluctuations in nutrient availability seems to be responsible for the cessation of reef corals. The sea level fall reached its climax in the subaerial exposure and palaeokarst development of the florigemma-Bank. The reef building corals are typical pioneer species. The blade-like, flattened F. amchnoides colonies are characterized by their light porous calcium carbonate skeleton, which is a distinct advantage in soft bottom environment. Thus, they settled on soft bottom exposing the large parts of its surface to the incoming light. On the other hand, in response to their light requirements they were also able to settle shaded canopy structures or reef caves. Th. dendroidea is an opportunistic coral species in very shallow, well illuminated marine environment. Their thin and densely spaced branches led to a very high surface/volume ratio of the colonies that were capable to exploit incoming light due to their small thamasterioid calices characterized by “highly integrated polyps”. In addition, sideward coalescence of branches during colony growth led to a wave-resistant framework and favoured the authochthonous preservation of the reefs. Asexual reproduction by fragmented colonies promoted reef development as Th. dendroidea thickets laterally extend over the sea floor or new reefs have developed from broken fragments of parent colonies. Similar build ups with Th. dendroidea as a dominant or frequent reef building coral species are known from the Paris Basin and elsewhere from the Lower Saxony Basin (Kleiner Deister Mts). These buildups developed in well-illuminated shallow water and encompass coral reefs or coral thrombolite reefs. Intra- and inter-reef deposits vary between well-winnowed reef debris limestone and mudstones representing considerably calmer conditions. Solenoporid, nerineids and diceratides belong to the characteristic fossils of these occurrences. However, diceratides are missing in theflorigemma-Bank Member. Th. dendroidea differs in its colonization of low- to high-energy environment from recent ramose scleractinian corals (e.g., Acropora and Porites sp.). The latter are restricted to agitated water habitats creating coral thickets and carpets. According to the morphologic plasticity of Th. dendroidea, thick-branched colonies developed in a milieu of high water energy, whereas fragile, wide- and thin-branched colonies prevail in low-energy settings. Due to its relatively rapid growth, Th. dendroidea was able to keep pace with increased Sedimentation rates. 68 benthonic foraminiferan species/taxa have been recognized in thin sections. Agglutinated foraminifers (textulariids) predominate when compared with rotaliids and milioliids. Numerous species are restricted to a certain facies type or occur in higher population densities, in particular Everticyclammina sp., a larger agglutinated foraminifer that occurs in rock building amounts. Among the 25 reef dwelling foraminiferal species, a few were so far only known from Late Jurassic sponge reefs. Another striking feature is the frequency of adherent foraminiferal species. Fauna and flora, in particular dasycladaleans and agglutinated foraminifers, document palaeobiogeographic relationships to the Tethys and point to (sub)tropical conditions. Moreover, in Germany this foraminiferan assemblage is yet uncompared. In Southern Germany similar tethyan type assemblages are not present in strata as young as Middle Tithonian.

Rôle de la conformation des glycoprotéines de l’enveloppe du VIH-1 dans la réponse cytotoxique cellulaire dépendante des anticorps et impact des protéines virales Nef et Vpu

Alors que d’énormes efforts sont mis de l’avant pour mettre en place des stratégies thérapeutiques contre l’infection au VIH-1, il est nécessaire de mieux cerner les déterminants viraux qui aideront à l’efficacité de celles-ci. En ce sens, une volumineuse littérature scientifique suggère que les anticorps contre le VIH-1 possédant une capacité à induire une réponse effectrice dépendante de leur portion Fc puissent jouer un rôle important dans la prévention de l’infection et dans la progression de la maladie. Cependant, peu d’information est disponible concernant les déterminants reconnus par ces anticorps et comment le virus s’en protège. Le but des travaux présentés dans cette thèse est donc d’élucider les mécanismes viraux contrôlant la reconnaissance des cellules infectées par ces anticorps capables d’induire une réponse effectrice. De par les corrélats de protection identifiés au cours de l’essai vaccinal RV144, les travaux présentés ici se concentrent sur la réponse cytotoxique dépendante des anticorps (ADCC), puisqu’il s’agit d’une réponse effectrice suggérée pour avoir joué un rôle dans la protection observée dans le RV144, seul essai vaccinal anti-VIH à avoir démontré un certain degré de protection. De plus, plusieurs anticorps capables d’induire cette réponse contre le VIH sont connus pour reconnaître les glycoprotéines de surface du virus (Env) dans une conformation dite ouverte, c’est-à-dire la conformation adoptée lors de la liaison d’Env avec son récepteur CD4 (épitopes CD4i). Nous avons mis au point deux techniques in vitro permettant d’étudier ces changements de conformation ainsi que leur impact sur la réponse ADCC. Les techniques mises au point, un ÉLISA sur base cellulaire pour mesurer les changements de conformation d’Env ainsi que la mesure de la réponse ADCC par cytométrie en flux, nous ont permis de démontrer comment le virus empêche l’exposition des épitopes d’Env CD4i. L’activité simultanée des protéines accessoires virales Nef et Vpu sur le retrait du récepteur CD4 de la surface des cellules infectées et l’inhibition du facteur de restriction Tétherine / BST-2 par Vpu contrôlent à la fois les niveaux d’Env et de CD4 à la surface cellulaire et donc modulent l’interaction Env-CD4 et ultimement la susceptibilité à la réponse ADCC contre les épitopes CD4i reconnus par des anticorps hautement prévalents lors de l’infection au VIH. Également, nous démontrons comment de petits composés mimant la liaison de CD4 sur Env sont capables de forcer l’exposition des épitopes CD4i, même en présence des protéines Nef et Vpu, et donc d’augmenter la susceptibilité des cellules infectées à la réponse ADCC. Une autre découverte présentée ici est la démonstration que la portion soluble d’Env produite par les cellules infectées peut interagir avec le récepteur CD4 des cellules non-infectées avoisinantes et induire leur reconnaissance et élimination par la réponse ADCC contre Env. Somme toute, la modulation de la réponse ADCC par l’interaction Env–CD4 représente un important pilier de la relation hôte – pathogène du VIH-1 de la perspective des réponses Fc-dépendantes. Les travaux présentés dans cette thèse ont le potentiel d’être utilisés dans l’élaboration de nouvelles stratégies antivirales tout en élargissant les connaissances fondamentales de cette interaction hôte – pathogène.

Model Probability in Self-organising Maps

Growing models have been widely used for clustering or topology learning. Traditionally these models work on stationary environments, grow incrementally and adapt their nodes to a given distribution based on global parameters. In this paper, we present an enhanced unsupervised self-organising network for the modelling of visual objects. We first develop a framework for building non-rigid shapes using the growth mechanism of the self-organising maps, and then we define an optimal number of nodes without overfitting or underfitting the network based on the knowledge obtained from information-theoretic considerations. We present experimental results for hands and we quantitatively evaluate the matching capabilities of the proposed method with the topographic product.

Design, Fabrication, and Mechanical Property Analysis of 3D Nanoarchitected Materials

Recent developments in micro- and nanoscale 3D fabrication techniques have enabled the creation of materials with a controllable nanoarchitecture that can have structural features spanning 5 orders of magnitude from tens of nanometers to millimeters. These fabrication methods in conjunction with nanomaterial processing techniques permit a nearly unbounded design space through which new combinations of nanomaterials and architecture can be realized. In the course of this work, we designed, fabricated, and mechanically analyzed a wide range of nanoarchitected materials in the form of nanolattices made from polymer, composite, and hollow ceramic beams. Using a combination of two-photon lithography and atomic layer deposition, we fabricated samples with periodic and hierarchical architectures spanning densities over 4 orders of magnitude from ρ=0.3-300kg/m³ and with features as small as 5nm. Uniaxial compression and cyclic loading tests performed on different nanolattice topologies revealed a range of novel mechanical properties: the constituent nanoceramics used here have size-enhanced strengths that approach the theoretical limit of materials strength; hollow aluminum oxide (Al₂O₃) nanolattices exhibited ductile-like deformation and recovered nearly completely after compression to 50% strain when their wall thicknesses were reduced below 20nm due to the activation of shell buckling; hierarchical nanolattices exhibited enhanced recoverability and a near linear scaling of strength and stiffness with relative density, with E∝ρ^1.04 and σy∝ρ^1.17 for hollow Al₂O₃ samples; periodic rigid and non-rigid nanolattice topologies were tested and showed a nearly uniform scaling of strength and stiffness with relative density, marking a significant deviation from traditional theories on “bending” and “stretching” dominated cellular solids; and the mechanical behavior across all topologies was highly tunable and was observed to strongly correlate with the slenderness λ and the wall thickness-to-radius ratio t/a of the beams. These results demonstrate the potential of nanoarchitected materials to create new highly tunable mechanical metamaterials with previously unattainable properties.

Análise de imagens em Geologia

O curso proposto está dividido em sete capítulos que vão desde a apresentação da importância da análise de imagens em geologia até à discussão e aplicação de aprendizagem máquina na análise de imagens. Sou defensor do software livre, assim todos os programas utilizados neste curso caiem nesta categoria. Os exemplos apresentados serão demonstrados com recurso aos seguintes programas:  QGIS – Sistemas de informação geográfica  GIMP – Tratamento de imagens  R - Cálculo  RStudio – IDE para o R  Anaconda Python Notebook – IDE para Python  OpenCV – Visão computacional Pretendo que o curso para o qual este texto serve de suporte seja eminentemente prático, um curso de “mãos na massa”, esperando-se que cada participante possa tratar temas do seu interesse pessoal. No primeiro capítulo é feita uma introdução sobre o que são imagens e a sua importância em geologia. O segundo capítulo trata de descrever os passos para a instalação do software proposto e fornecer pequenos exemplos da sua utilização. O terceiro capítulo descreve os métodos e as limitações da aquisição das imagens. São dados alguns exemplos de funções de aquisição de imagens. Os exemplos práticos deste capítulo incluem exemplos em Python e R. O quarto capítulo fala dos parâmetros contidos num ficheiro de imagens. Neste capítulo são apresentados exemplos em Python. O quinto capítulo trata das ferramentas que se podem aplicar durante o préprocessamento de uma imagem. O sexto capítulo trata de mostrar alguns exemplos de análise de imagens e no sétimo capítulo é abordada a questão de utilização de algoritmos de aprendizagem máquina na análise de imagens.

Non-evolutive pattern recognition techniques: An application in medical image diagnostics

Lo studio dell’intelligenza artificiale si pone come obiettivo la risoluzione di una classe di problemi che richiedono processi cognitivi difficilmente codificabili in un algoritmo per essere risolti. Il riconoscimento visivo di forme e figure, l’interpretazione di suoni, i giochi a conoscenza incompleta, fanno capo alla capacità umana di interpretare input parziali come se fossero completi, e di agire di conseguenza. Nel primo capitolo della presente tesi sarà costruito un semplice formalismo matematico per descrivere l’atto di compiere scelte. Il processo di “apprendimento” verrà descritto in termini della massimizzazione di una funzione di prestazione su di uno spazio di parametri per un ansatz di una funzione da uno spazio vettoriale ad un insieme finito e discreto di scelte, tramite un set di addestramento che descrive degli esempi di scelte corrette da riprodurre. Saranno analizzate, alla luce di questo formalismo, alcune delle più diffuse tecniche di artificial intelligence, e saranno evidenziate alcune problematiche derivanti dall’uso di queste tecniche. Nel secondo capitolo lo stesso formalismo verrà applicato ad una ridefinizione meno intuitiva ma più funzionale di funzione di prestazione che permetterà, per un ansatz lineare, la formulazione esplicita di un set di equazioni nelle componenti del vettore nello spazio dei parametri che individua il massimo assoluto della funzione di prestazione. La soluzione di questo set di equazioni sarà trattata grazie al teorema delle contrazioni. Una naturale generalizzazione polinomiale verrà inoltre mostrata. Nel terzo capitolo verranno studiati più nel dettaglio alcuni esempi a cui quanto ricavato nel secondo capitolo può essere applicato. Verrà introdotto il concetto di grado intrinseco di un problema. Verranno inoltre discusse alcuni accorgimenti prestazionali, quali l’eliminazione degli zeri, la precomputazione analitica, il fingerprinting e il riordino delle componenti per lo sviluppo parziale di prodotti scalari ad alta dimensionalità. Verranno infine introdotti i problemi a scelta unica, ossia quella classe di problemi per cui è possibile disporre di un set di addestramento solo per una scelta. Nel quarto capitolo verrà discusso più in dettaglio un esempio di applicazione nel campo della diagnostica medica per immagini, in particolare verrà trattato il problema della computer aided detection per il rilevamento di microcalcificazioni nelle mammografie.

Spatio-temporal alignment of pedobarographic image sequences

O documento em anexo encontra-se na versão post-print (versão corrigida pelo editor).

Global localization with non-quantized local image features

In the field of appearance-based robot localization, the mainstream approach uses a quantized representation of local image features. An alternative strategy is the exploitation of raw feature descriptors, thus avoiding approximations due to quantization. In this work, the quantized and non-quantized representations are compared with respect to their discriminativity, in the context of the robot global localization problem. Having demonstrated the advantages of the non-quantized representation, the paper proposes mechanisms to reduce the computational burden this approach would carry, when applied in its simplest form. This reduction is achieved through a hierarchical strategy which gradually discards candidate locations and by exploring two simplifying assumptions about the training data. The potential of the non-quantized representation is exploited by resorting to the entropy-discriminativity relation. The idea behind this approach is that the non-quantized representation facilitates the assessment of the distinctiveness of features, through the entropy measure. Building on this finding, the robustness of the localization system is enhanced by modulating the importance of features according to the entropy measure. Experimental results support the effectiveness of this approach, as well as the validity of the proposed computation reduction methods.

Algorithms for on-line image registration from multiple views

Images of a scene, static or dynamic, are generally acquired at different epochs from different viewpoints. They potentially gather information about the whole scene and its relative motion with respect to the acquisition device. Data from different (in the spatial or temporal domain) visual sources can be fused together to provide a unique consistent representation of the whole scene, even recovering the third dimension, permitting a more complete understanding of the scene content. Moreover, the pose of the acquisition device can be achieved by estimating the relative motion parameters linking different views, thus providing localization information for automatic guidance purposes. Image registration is based on the use of pattern recognition techniques to match among corresponding parts of different views of the acquired scene. Depending on hypotheses or prior information about the sensor model, the motion model and/or the scene model, this information can be used to estimate global or local geometrical mapping functions between different images or different parts of them. These mapping functions contain relative motion parameters between the scene and the sensor(s) and can be used to integrate accordingly informations coming from the different sources to build a wider or even augmented representation of the scene. Accordingly, for their scene reconstruction and pose estimation capabilities, nowadays image registration techniques from multiple views are increasingly stirring up the interest of the scientific and industrial community. Depending on the applicative domain, accuracy, robustness, and computational payload of the algorithms represent important issues to be addressed and generally a trade-off among them has to be reached. Moreover, on-line performance is desirable in order to guarantee the direct interaction of the vision device with human actors or control systems. This thesis follows a general research approach to cope with these issues, almost independently from the scene content, under the constraint of rigid motions. This approach has been motivated by the portability to very different domains as a very desirable property to achieve. A general image registration approach suitable for on-line applications has been devised and assessed through two challenging case studies in different applicative domains. The first case study regards scene reconstruction through on-line mosaicing of optical microscopy cell images acquired with non automated equipment, while moving manually the microscope holder. By registering the images the field of view of the microscope can be widened, preserving the resolution while reconstructing the whole cell culture and permitting the microscopist to interactively explore the cell culture. In the second case study, the registration of terrestrial satellite images acquired by a camera integral with the satellite is utilized to estimate its three-dimensional orientation from visual data, for automatic guidance purposes. Critical aspects of these applications are emphasized and the choices adopted are motivated accordingly. Results are discussed in view of promising future developments.

An Image Recognition Software for Identification of Vehicle Homologation Tags

In recent years, we have witnessed great changes in the industrial environment as a result of the innovations introduced by Industry 4.0, especially in the integration of Internet of Things, Automation and Robotics in the manufacturing field. The project presented in this thesis lies within this innovation context and describes the implementation of an Image Recognition application focused on the automotive field. The project aims at helping the supply chain operator to perform an effective and efficient check of the homologation tags present on vehicles. The user contribution consists in taking a picture of the tag and the application will automatically, exploiting Amazon Web Services, return the result of the control about the correctness of the tag, the correct positioning within the vehicle and the presence of faults or defects on the tag. To implement this application we ombined two IoT platforms widely used in industrial field: Amazon Web Services(AWS) and ThingWorx. AWS exploits Convolutional Neural Networks to perform Text Detection and Image Recognition, while PTC ThingWorx manages the user interface and the data manipulation.