Comparison of Local Feature Detectors and Descriptors for Visual Object Categorization

Local features are used in many computer vision tasks including visual object categorization, content-based image retrieval and object recognition to mention a few. Local features are points, blobs or regions in images that are extracted using a local feature detector. To make use of extracted local features the localized interest points are described using a local feature descriptor. A descriptor histogram vector is a compact representation of an image and can be used for searching and matching images in databases. In this thesis the performance of local feature detectors and descriptors is evaluated for object class detection task. Features are extracted from image samples belonging to several object classes. Matching features are then searched using random image pairs of a same class. The goal of this thesis is to find out what are the best detector and descriptor methods for such task in terms of detector repeatability and descriptor matching rate.

Automated, adaptive methods for forest inventory

Forest inventories are used to estimate forest characteristics and the condition of forest for many different applications: operational tree logging for forest industry, forest health state estimation, carbon balance estimation, land-cover and land use analysis in order to avoid forest degradation etc. Recent inventory methods are strongly based on remote sensing data combined with field sample measurements, which are used to define estimates covering the whole area of interest. Remote sensing data from satellites, aerial photographs or aerial laser scannings are used, depending on the scale of inventory. To be applicable in operational use, forest inventory methods need to be easily adjusted to local conditions of the study area at hand. All the data handling and parameter tuning should be objective and automated as much as possible. The methods also need to be robust when applied to different forest types. Since there generally are no extensive direct physical models connecting the remote sensing data from different sources to the forest parameters that are estimated, mathematical estimation models are of "black-box" type, connecting the independent auxiliary data to dependent response data with linear or nonlinear arbitrary models. To avoid redundant complexity and over-fitting of the model, which is based on up to hundreds of possibly collinear variables extracted from the auxiliary data, variable selection is needed. To connect the auxiliary data to the inventory parameters that are estimated, field work must be performed. In larger study areas with dense forests, field work is expensive, and should therefore be minimized. To get cost-efficient inventories, field work could partly be replaced with information from formerly measured sites, databases. The work in this thesis is devoted to the development of automated, adaptive computation methods for aerial forest inventory. The mathematical model parameter definition steps are automated, and the cost-efficiency is improved by setting up a procedure that utilizes databases in the estimation of new area characteristics.

Comparison measures of maps generated by geostatistical methods

This study uses several measures derived from the error matrix for comparing two thematic maps generated with the same sample set. The reference map was generated with all the sample elements and the map set as the model was generated without the two points detected as influential by the analysis of local influence diagnostics. The data analyzed refer to the wheat productivity in an agricultural area of 13.55 ha considering a sampling grid of 50 x 50 m comprising 50 georeferenced sample elements. The comparison measures derived from the error matrix indicated that despite some similarity on the maps, they are different. The difference between the estimated production by the reference map and the actual production was of 350 kilograms. The same difference calculated with the mode map was of 50 kilograms, indicating that the study of influential points is of fundamental importance to obtain a more reliable estimative and use of measures obtained from the error matrix is a good option to make comparisons between thematic maps.

Local control of human papillomavirus infection after anal condylomata acuminata eradication

OBJECTIVE: To verify whether the eradication of anal condylomata acuminata was effective for local control of HPV infection using anal colposcopy and anal brush cytology.METHODS: We evaluated 147 patients treated for anal margin and/or anal canal condyloma, with 108 HIV-positive and 39 HIV-negative individuals. The average age for males was 40 years for HIV-positive and 27.5 for HIV-negative. In females, the mean age was 37.5 years for HIV-positive and 31.5 for HIV-negative.RESULTS: Twenty-four patients (16.3%) had normal cytology and anal colposcopy, 16 (10.9%) normal cytology and altered anal colposcopy, 52 (35.4%) normal anal colposcopy and altered cytology, and 55 (37.4%) had altered cytology and anal colposcopy.CONCLUSION: the eradication of clinical lesions failed to locally control HPV infection.

Use of local muscle flaps to cover leg bone exposures

Objective: To evaluate the use of the medial gastrocnemius muscle and/or soleus muscle flaps as surgical treatment of the leg bone exposure.Methods: We retrospectively analyzed the medical records of patients undergoing transposition of the medial gastrocnemius and / or soleus for treating exposed bone in the leg, from January 1976 to July 2009, gathering information on epidemiological data, the etiology the lesion, the time between the initial injury and muscle transposition, the muscle used to cover the lesion, the healing evolution of the skin coverage and the function of the gastrocnemius-soleus unit.Results: 53 patients were operated, the ages varying between nine and 84 years (mean age 41); 42 were male and 11 female. The main initial injury was trauma (84.8%), consisting of tibia and / or fibula fracture. The most frequently used muscle was the soleus, in 40 cases (75.5%). The rank of 49 patients (92.5%) was excellent or good outcome, of three (5.6%) as regular and of one (1.9%) as unsatisfactory.Conclusion: the treatment of bone exposure with local muscle flaps (gastrocnemius and/or soleus) enables obtaining satisfactory results in covering of exposed structures, favoring local vascularization and improving the initial injury. It offers the advantage of providing a treatment in only one surgical procedure, an earlier recovery and reduced hospital stay.

Assesment of the treatment of earlobe keloids with triamcinolone injections, surgical resection, and local pressure

OBJECTIVE: To evaluate the combined treatment of ear lobe keloids. METHODS: We studied 46 consecutive patients with 81 ear lobe keloids. Patients underwent local infiltration of triamcinolone acetonide (TCN) at concentrations of 40mg/ml (Group 1), 20 mg/ml (Group 2) and 10mg/ml (Group 3). The volume of TCN infiltrate varied according to the size of the lesion. Treatment consisted of three monthly injections before surgery, excision of keloid in the fourth month and perioperative infiltration, followed by two more leaks TCN within two months. Patients used earrings pressure on the scar after operation for four months. The pressure exerted by earrings in the ear lobe was measured electronically. Post-treatment follow-up of patients was 24 months. RESULTS: TCN at concentrations of 20mg/ml and 40mg/ml were effective for the treatment of keloids, no difference between the groups (p = 0.58). However, patients in which TCN was infiltrated the 10mg/ml had poor involution of keloid and the study of this group was stopped. CONCLUSION: the combination of infiltration TCN month to 20 mg/mL (1.2mg to 2.0mg per mm3 TCN injury), surgical excision and pressure application device is effective for treatment of keloid ear lobe.

Local Features in Image and Video Processing - Object Class Matching and Video Shot Detection

The usage of digital content, such as video clips and images, has increased dramatically during the last decade. Local image features have been applied increasingly in various image and video retrieval applications. This thesis evaluates local features and applies them to image and video processing tasks. The results of the study show that 1) the performance of different local feature detector and descriptor methods vary significantly in object class matching, 2) local features can be applied in image alignment with superior results against the state-of-the-art, 3) the local feature based shot boundary detection method produces promising results, and 4) the local feature based hierarchical video summarization method shows promising new new research direction. In conclusion, this thesis presents the local features as a powerful tool in many applications and the imminent future work should concentrate on improving the quality of the local features.

Application and development of numerical methods for the modelling of innovative gas cooled fission reactors

Innovative gas cooled reactors, such as the pebble bed reactor (PBR) and the gas cooled fast reactor (GFR) offer higher efficiency and new application areas for nuclear energy. Numerical methods were applied and developed to analyse the specific features of these reactor types with fully three dimensional calculation models. In the first part of this thesis, discrete element method (DEM) was used for a physically realistic modelling of the packing of fuel pebbles in PBR geometries and methods were developed for utilising the DEM results in subsequent reactor physics and thermal-hydraulics calculations. In the second part, the flow and heat transfer for a single gas cooled fuel rod of a GFR were investigated with computational fluid dynamics (CFD) methods. An in-house DEM implementation was validated and used for packing simulations, in which the effect of several parameters on the resulting average packing density was investigated. The restitution coefficient was found out to have the most significant effect. The results can be utilised in further work to obtain a pebble bed with a specific packing density. The packing structures of selected pebble beds were also analysed in detail and local variations in the packing density were observed, which should be taken into account especially in the reactor core thermal-hydraulic analyses. Two open source DEM codes were used to produce stochastic pebble bed configurations to add realism and improve the accuracy of criticality calculations performed with the Monte Carlo reactor physics code Serpent. Russian ASTRA criticality experiments were calculated. Pebble beds corresponding to the experimental specifications within measurement uncertainties were produced in DEM simulations and successfully exported into the subsequent reactor physics analysis. With the developed approach, two typical issues in Monte Carlo reactor physics calculations of pebble bed geometries were avoided. A novel method was developed and implemented as a MATLAB code to calculate porosities in the cells of a CFD calculation mesh constructed over a pebble bed obtained from DEM simulations. The code was further developed to distribute power and temperature data accurately between discrete based reactor physics and continuum based thermal-hydraulics models to enable coupled reactor core calculations. The developed method was also found useful for analysing sphere packings in general. CFD calculations were performed to investigate the pressure losses and heat transfer in three dimensional air cooled smooth and rib roughened rod geometries, housed inside a hexagonal flow channel representing a sub-channel of a single fuel rod of a GFR. The CFD geometry represented the test section of the L-STAR experimental facility at Karlsruhe Institute of Technology and the calculation results were compared to the corresponding experimental results. Knowledge was gained of the adequacy of various turbulence models and of the modelling requirements and issues related to the specific application. The obtained pressure loss results were in a relatively good agreement with the experimental data. Heat transfer in the smooth rod geometry was somewhat under predicted, which can partly be explained by unaccounted heat losses and uncertainties. In the rib roughened geometry heat transfer was severely under predicted by the used realisable k − epsilon turbulence model. An additional calculation with a v2 − f turbulence model showed significant improvement in the heat transfer results, which is most likely due to the better performance of the model in separated flow problems. Further investigations are suggested before using CFD to make conclusions of the heat transfer performance of rib roughened GFR fuel rod geometries. It is suggested that the viewpoints of numerical modelling are included in the planning of experiments to ease the challenging model construction and simulations and to avoid introducing additional sources of uncertainties. To facilitate the use of advanced calculation approaches, multi-physical aspects in experiments should also be considered and documented in a reasonable detail.

Local Binary Patterns in Focal-Plane Processing. Analysis and Applications

Feature extraction is the part of pattern recognition, where the sensor data is transformed into a more suitable form for the machine to interpret. The purpose of this step is also to reduce the amount of information passed to the next stages of the system, and to preserve the essential information in the view of discriminating the data into different classes. For instance, in the case of image analysis the actual image intensities are vulnerable to various environmental effects, such as lighting changes and the feature extraction can be used as means for detecting features, which are invariant to certain types of illumination changes. Finally, classification tries to make decisions based on the previously transformed data. The main focus of this thesis is on developing new methods for the embedded feature extraction based on local non-parametric image descriptors. Also, feature analysis is carried out for the selected image features. Low-level Local Binary Pattern (LBP) based features are in a main role in the analysis. In the embedded domain, the pattern recognition system must usually meet strict performance constraints, such as high speed, compact size and low power consumption. The characteristics of the final system can be seen as a trade-off between these metrics, which is largely affected by the decisions made during the implementation phase. The implementation alternatives of the LBP based feature extraction are explored in the embedded domain in the context of focal-plane vision processors. In particular, the thesis demonstrates the LBP extraction with MIPA4k massively parallel focal-plane processor IC. Also higher level processing is incorporated to this framework, by means of a framework for implementing a single chip face recognition system. Furthermore, a new method for determining optical flow based on LBPs, designed in particular to the embedded domain is presented. Inspired by some of the principles observed through the feature analysis of the Local Binary Patterns, an extension to the well known non-parametric rank transform is proposed, and its performance is evaluated in face recognition experiments with a standard dataset. Finally, an a priori model where the LBPs are seen as combinations of n-tuples is also presented

New methods for controlling twin configurations and characterizing twin boundaries in 5M Ni-Mn-Ga for the development of applications

Traditional methods for studying the magnetic shape memory (MSM) alloys Ni-Mn-Ga include subjecting the entire sample to a uniform magnetic field or completely actuating the sample mechanically. These methods have produced significant results in characterizing the MSM effect, the properties of Ni-Mn-Ga and have pioneered the development of applications from this material. Twin boundaries and their configuration within a Ni-Mn-Ga sample are a key component in the magnetic shape memory effect. Applications that are developed require an understanding of twin boundary characteristics and, more importantly, the ability to predictably control them. Twins have such a critical role that the twinning stress of a Ni-Mn-Ga crystal is the defining characteristic that indicates its quality and significant research has been conducted to minimize this property. This dissertation reports a decrease in the twinning stress, predictably controlling the twin configuration and characterizing the dynamics of twin boundaries. A reduction of the twinning stress is demonstrated by the discovery of Type II twins within Ni-Mn-Ga which have as little as 10% of the twinning stress of traditional Type I twins. Furthermore, new methods of actuating a Ni-Mn-Ga element using localized unidirectional or bidirectional magnetic fields were developed that can predictably control the twin configuration in a localized area of a Ni-Mn-Ga element. This method of controlling the local twin configuration was used in the characterization of twin boundary dynamics. Using a localized magnetic pulse, the velocity and acceleration of a single twin boundary were measured to be 82.5 m/s and 2.9 × 107 m/s2, and the time needed for the twin boundary to nucleate and begin moving was less than 2.8 μs. Using a bidirectional magnetic field from a diametrically magnetized cylindrical magnet, a highly reproducible and controllable local twin configuration was created in a Ni-Mn-Ga element which is the fundamental pumping mechanism in the MSM micropump that has been co-invented and extensively characterized by the author.

Thick Section Laser Beam Welding of Structural Steels: Methods for Improving Welding Efficiency

Laser beam welding (LBW) is applicable for a wide range of industrial sectors and has a history of fifty years. However, it is considered an unusual method with applications typically limited to welding of thin sheet metal. With a new generation of high power lasers there has been a renewed interest in thick section LBW (also known as keyhole laser welding). There was a growing body of publications during 2001-2011 that indicates an increasing interest in laser welding for many industrial applications, and in last ten years, an increasing number of studies have examined the ways to increase the efficiency of the process. Expanding the thickness range and efficiency of LBW makes the process a possibility for industrial applications dealing with thick metal welding: shipbuilding, offshore structures, pipelines, power plants and other industries. The advantages provided by LBW, such as high process speed, high productivity, and low heat input, may revolutionize these industries and significantly reduce the process costs. The research to date has focused on either increasing the efficiency via optimizing process parameters, or on the process fundamentals, rather than on process and workpiece modifications. The argument of this thesis is that the efficiency of the laser beam process can be increased in a straightforward way in the workshop conditions. Throughout this dissertation, the term “efficiency” is used to refer to welding process efficiency, specifically, an increase in efficiency refers an increase in weld’s penetration depth without increasing laser power level or decreasing welding speed. These methods are: modifications of the workpiece – edge surface roughness and air gap between the joining plates; modification of the ambient conditions – local reduction of the pressure in the welding zone; modification of the welding process – preheating of the welding zone. Approaches to improve the efficiency are analyzed and compared both separately and combined. These experimentally proven methods confirm previous findings and contribute additional evidence which expand the opportunities for laser beam welding applications. The focus of this research was primarily on the effects of edge surface roughness preparation and pre-set air gap between the plates on weld quality and penetration depth. To date, there has been no reliable evidence that such modifications of the workpiece give a positive effect on the welding efficiency. Other methods were tested in combination with the two methods mentioned above. The most promising - combining with reduced pressure method - resulted in at least 100% increase in efficiency. The results of this thesis support the idea that joining those methods in one modified process will provide the modern engineering with a sufficient tool for many novel applications with potential benefits to a range of industries.

Effects of peeling methods on the quality of cubiu fruits

Cubiu (Solanum sessiliflorum Dunal) is an Amazonian Basin native fruit. Its importance comes from its high contents of pectin. Currently, processing technologies are necessary for the substitution of the traditional system (small crops and small-scale processing) for a larger scale system and thus increase the use of biodiversity and promote the implementation of Local Productive Arrangements of agribusiness in the Amazon. This research aims to evaluate the methods of peeling cubiu. Ripe fruits were divided into lots (150 each) and subjected to the following treatments: immersion in 2.5% NaOH boiling solution for 5 minutes, exposure to water vapor, and immersion in water at 96 ºC for 5, 10, 15 and 20 minutes. The peel released during heat treatment and immediately removed under running tap water. In the control treatment, the fruits were manually peeled (unheated) with a stainless steel knife. The treatments were evaluated for completeness and ease of peeling, tissue integrity, texture, and peroxidase activity. The immersion in 2.5% NaOH boiling solution (5 minutes) stood out as the best treatment since it inhibited the enzymatic browning and intensified the natural yellow color of the cubiu fruit and easily and fully peeled the whole fruit more rapidly without damaging its tissues. This treatment was chosen as the most advantageous because it can promote simultaneous peeling and bleaching. Therefore, it is recommended for cubiu industrial processing.

Determination of the thermal parameters of high-power batteries by local heat

A new approach to the determination of the thermal parameters of high-power batteries is introduced here. Application of local heat flux measurement with a gradient heat flux sensor (GHFS) allows determination of the cell thermal parameters in di_erent surface points of the cell. The suggested methodology is not cell destructive as it does not require deep discharge of the cell or application of any charge/discharge cycles during measurements of the thermal parameters of the cell. The complete procedure is demonstrated on a high-power Li-ion pouch cell, and it is verified on a sample with well-known thermal parameters. A comparison of the experimental results with conventional thermal characterization methods shows an acceptably low error. The dependence of the cell thermal parameters on state of charge (SoC) and measurement points on the surface was studied by the proposed measurement approach.

Reactive Hyperemia as endothelial function determinant using plethysmography methods

L’atteinte de la fonction endothéliale représente une phase précoce de l’athérosclérose, un stade où les patients sont généralement asymptomatiques. Il existe donc un intérêt certain à détecter la dysfonction endothéliale. Nous avons développé une technique de mesure des variations de flot artériel au niveau des membres supérieurs, basée sur la spectroscopie proche infrarouge (NIRS). Cette approche permettrait d’étudier le niveau d’atteinte vasculaire et probablement de quantifier le degré de dysfonction endothéliale périphérique lors d’une hyperémie réactive. L'expérience a été exécutée sur deux cohortes de 13 et de 15 patients et a été comparée à la pléthysmographie par jauge de contrainte (SGP) qui est considérée comme une méthode de référence. Par la suite, nous avons caractérisé la réponse endothéliale par modélisation de la courbe hyperémique du flot artériel. Des études préliminaires avaient démontré que la réponse hyperémique adoptait majoritairement une forme bi-modale. Nous avons tenté de séparer les composantes endothéliales-dépendantes et endothéliales-indépendantes de l’hyperémie. La quantification des deux composantes de la réaction hyperémique permet de calculer un indice de la ‘santé’ du système endothélial local. Cet indice est nommé le ηfactor. Les résultats montrent une forte corrélation des mesures de flots entre la technique développée et la méthode de référence (r=0.91). Nous avons conclu que NIRS est une approche précise pour la mesure non-invasive du flot artériel. Nous avons obtenu une bonne répétabilité (ICC = 0.9313) pour le ηfactor indiquant sa robustesse. Cependant des études supplémentaires sont nécessaires pour valider la valeur de diagnostic du facteur défini. Mots clés: hyperémie réactive, réponse myogénique, oxyde nitrique, athérosclérose, spectroscopie proche infrarouge

Dieu, le capitalisme et le développement local : conflits sociaux et enracinement territorial : étude monographique d'un village québécois

Inscrite dans la tradition monographique en sociologie et en anthropologie, cette thèse prend pour objet la diversité des pratiques et des idéologies caractéristiques des différents types de populations rurales distingués en fonction de l'enracinement territorial, afin d'apporter un nouvel éclairage sur les conflits sociaux actuels dans tous les milieux ruraux québécois qui surgissent notamment de l'accroissement du nombre des néo-ruraux dont les visions du monde s'opposent à celles des agriculteurs, dont le nombre diminue sans cesse. Prenant un village comme observatoire, il s'agit de rendre compte du mouvement totalisant de l'expérience de la vie en société à la fois dans ses dimensions « matérielles » et « symboliques ». L'étude des principales formes de vie sociale (religieuse, économique et politique) se fait grâce à des méthodes diversifiées: l'observation participante, l'analyse statistique, l'analyse du discours, le travail sur les archives municipales et l'histoire orale. L'analyse des différentes formes de vie sociale montre que leur organisation est structurée par deux principaux modèles. Le modèle public et communautaire comprend les personnes qui valorisent l'implication de l'État et des professionnels dans la gestion collective de la redistribution des richesses et dans le développement des milieux ruraux. Ces personnes occupent une position économique « marginale » à l'intérieur de la localité et sont plus près des milieux urbains tant par leurs positions que par leurs visions du monde. Quant au modèle privé et familial, il comprend les personnes défendant le rôle prépondérant des réseaux familiaux dans le développement local et la fermeture de la localité face à la concurrence des marchés extérieurs et aux interventions politiques exogènes. Les représentants de ce modèle occupent une position économique locale dominante, mais se sentent de plus en plus dominés politiquement face aux interventions extérieures des représentants politiques régionaux et des professionnels ainsi qu'économiquement à l'échelle mondiale où ils occupent une position dominée. Les oppositions sous-jacentes à ces deux modèles s'inscrivent dans une histoire ancienne qui met en scène d'une part les élites traditionnelles liées à l'Église et les notables francophones scolarisées et d'autre part les élites industrielles et commerciales qui succèdent aux anglophones dès les années 1920. Le sens et le contenu des modèles varient légèrement avec les transformations récentes de la structure familiale et la régionalisation des pouvoirs politiques et religieux.