Depuis ces dernières décennies, peu de sujets de la géologie de l’ingénieur ont attiré autant l’attention de la communauté scientifique que ceux traitant des risques naturels et plus particulièrement des instabilités gravitaires de grandes ampleurs reconnues sous les termes de DSGSD (Deep Seated Gravitational Slope Deformation) et de DSL (Deep Seated Landslide). Basée sur un nombre limité de cas d’étude à l’échelle naturelle (couplage datation et surveillance récente), la dynamique temporelle à court terme reste l’un de leurs aspects les moins étudiés aujourd’hui. Nous avons choisi l’exemple du massif de la Cristallère, situé en haute Vallée d’Aspe dans les Pyrénées béarnaises, où le DSL de la Cristallère a été récemment identifié et analysé à partir de deux méthodes de datation.Dans la continuité de ce travail, notre première approche consiste à évaluer la dynamique temporelle à court terme des mouvements de ce versant à différentes échelles et à partir de différentes méthodes de positionnement par satellites (constellations GPS et GLONASS) : positionnement multistations RGP (Réseau Géodésique Permanent), statique géodésique avec pivot et statique rapide avec pivot. Nous insistons dans ce travail, compte tenu des résultats originaux obtenus, sur l’intérêt du différentiel GNSS (Global Navigation Satellite System) avec un pivot qui se doit d’être proche géographiquement des observations à mener afin d’assurer une précision suffisante (de l’ordre du cm).Les observations et l'exploitation des modèles numériques de terrain issues des relevés aériens à différentes échelles obtenus avec un drone du type « aile volante » constituent la deuxième approche de ce travail de recherche. Elles viennent compléter la cartographie du site et démontrer l’existence d’un DSGSD à partir d’une caractérisation géomorphométrique haute résolution et haute précision (de l’ordre du cm également) ; elles permettent d’affiner la délimitation du DSL de la Cristallère et de sa zone la plus active (Pène du Thès) et du DSL du Peilhou.À ces deux approches complémentaires (combinaison de données tridimensionnelles ponctuelles et surfaciques), s’ajoutent préalablement un relevé de terrain géologique etgéophysique (mesures structurales, tomographies de résistivités électriques et profils électromagnétiques Very Low Frequency) ainsi qu’une analyse des archives disponibles sur les ouvrages présents dans le versant instable, tels que la galerie drainante de la centrale hydroélectrique du Baralet et l’ancien tunnel ferroviaire du Peilhou.L’utilisation combinée de ces trois approches confirme que les mouvements profonds du massif de la Cristallère sont encore actifs avec un forçage sismique avéré : la prise de conscience du rôle que peut jouer le DSGSD est tout aussi fondamentale. Ainsi tous les ouvrages existants dans le massif, ou les projets d’aménagement ou de réhabilitation sur ce site doivent considérer l’existence d’une déstructuration, lente et progressive de l’ensemble du massif (DSGSD). Par ailleurs, la méthodologie développée dans ce travail se veut générale et peut être utilisée pour surveiller et suivre, à court et moyen terme, tous types de mouvement de terrain, et en particulier les glissements de terrain ou rocheux, profonds ou superficiels, lents ou rapides. / For these last decades, few subjects of the geology of the engineer have drawn the attention of the scientific community as much as those dealing of the natural hazards and more particularly with large-scale gravitational instabilities known as DSGSD (Deep Seated Gravitational Slope Deformation) and DSL (Deep Seated Landslide). Based on few cases study on a natural scale (dating and recent monitoring), short term temporal dynamics remains one of their least studied aspects today. We made the choice of the Cristallère massif as an example. It is located in Upper Aspe Valley of the Pyrenees (Béarn region). The Cristallère DSL was recently identified and analyzed through two dating methods.On the basis of this work, our first approach consists in assessing short term temporal dynamics of these slopes movements on various scales and with various methods of satellites positioning (GPS and GLONASS constellations): multistation positioning RGP (Permanent Geodetic Network in France), statics geodetic with pivot and fast static with pivot. We insist in this work, given the original results obtained, on the interest of the differential GNSS (Global Navigation Satellite System) with a pivot which must be geographically close to the observations to be carried out in order to ensure sufficient accuracy (cm).Observations and exploitation of drone digital elevation models of the site from aerial surveys at different scales obtained with a “flying wing” are the second approach to this research work. They complete the mapping of the site and demonstrate the existence of a DSGSD based on a high resolution and high precision geomorphometric characterization (cm); they make it possible to refine the delimitation of the Cristallère DSL and its most active area (Pène du Thès) and the Peilhou DSL.In addition to these two complementary approaches, a geological and geophysical survey (structural geology measurements, electrical resistivity tomographies and electromagnetic profiles Very Low Frequency surveys) and an analysis of the available data on the structures present in the unstable slope, such as the large diameter underground water pipe for the hydroelectric plant of Baralet and the former railway tunnel of Peilhou. The combined use of these three approaches confirms that the deep movements of the Cristallère massif are still active with proven seismic forcing: awareness of the potential role taken by the DSGSD is then fundamental. Thus, all the existing structures in the massif or the projects on this site (new constructions or rehabilitation of old structures) have to consider the existence of a slow and gradual change of the entire massif (DSGSD). Moreover, the methodology developed in this work is intended to be general. It also makes possible to monitor and follow, in the short and medium term, all types of ground movement, in particular landslides or rockslides, deep or superficial, slow or fast.

A computação visual é uma área do conhecimento que estuda o desenvolvimento de sistemas artificiais capazes de detectar e desenvolver a percepção do meio ambiente através de informações de imagem ou dados multidimensionais. A percepção visual e a manipulação são combinadas em sistemas robóticos através de duas etapas "olhar"e depois "movimentar-se", gerando um laço de controle de feedback visual. Neste contexto, existe um interesse crescimente no uso dessas técnicas em veículos aéreos não tripulados (VANTs), também conhecidos como drones. Essas técnicas são aplicadas para posicionar o drone em modo de vôo autônomo, ou para realizar a detecção de regiões para vigilância aérea ou pontos de interesse. Os sistemas de computação visual geralmente tomam três passos em sua operação, que são: aquisição de dados em forma numérica, processamento de dados e análise de dados. A etapa de aquisição de dados é geralmente realizada por câmeras e sensores de proximidade. Após a aquisição de dados, o computador embarcado realiza o processamento de dados executando algoritmos com técnicas de medição (variáveis, índice e coeficientes), detecção (padrões, objetos ou áreas) ou monitoramento (pessoas, veículos ou animais). Os dados processados são analisados e convertidos em comandos de decisão para o controle para o sistema robótico autônomo Visando realizar a integração dos sistemas de computação visual com as diferentes plataformas de VANTs, este trabalho propõe o desenvolvimento de um framework para controle de missão e guiamento de VANTs baseado em visão computacional. O framework é responsável por gerenciar, codificar, decodificar e interpretar comandos trocados entre as controladoras de voo e os algoritmos de computação visual. Como estudo de caso, foram desenvolvidos dois algoritmos destinados à aplicação em agricultura de precisão. O primeiro algoritmo realiza o cálculo de um coeficiente de reflectância visando a aplicação auto-regulada e eficiente de agroquímicos, e o segundo realiza a identificação das linhas de plantas para realizar o guiamento dos VANTs sobre a plantação. O desempenho do framework e dos algoritmos propostos foi avaliado e comparado com o estado da arte, obtendo resultados satisfatórios na implementação no hardware embarcado. / Cumputer Vision is an area of knowledge that studies the development of artificial systems capable of detecting and developing the perception of the environment through image information or multidimensional data. Nowadays, vision systems are widely integrated into robotic systems. Visual perception and manipulation are combined in two steps "look" and then "move", generating a visual feedback control loop. In this context, there is a growing interest in using computer vision techniques in unmanned aerial vehicles (UAVs), also known as drones. These techniques are applied to position the drone in autonomous flight mode, or to perform the detection of regions for aerial surveillance or points of interest. Computer vision systems generally take three steps to the operation, which are: data acquisition in numerical form, data processing and data analysis. The data acquisition step is usually performed by cameras or proximity sensors. After data acquisition, the embedded computer performs data processing by performing algorithms with measurement techniques (variables, index and coefficients), detection (patterns, objects or area) or monitoring (people, vehicles or animals). The resulting processed data is analyzed and then converted into decision commands that serve as control inputs for the autonomous robotic system In order to integrate the visual computing systems with the different UAVs platforms, this work proposes the development of a framework for mission control and guidance of UAVs based on computer vision. The framework is responsible for managing, encoding, decoding, and interpreting commands exchanged between flight controllers and visual computing algorithms. As a case study, two algorithms were developed to provide autonomy to UAVs intended for application in precision agriculture. The first algorithm performs the calculation of a reflectance coefficient used to perform the punctual, self-regulated and efficient application of agrochemicals. The second algorithm performs the identification of crop lines to perform the guidance of the UAVs on the plantation. The performance of the proposed framework and proposed algorithms was evaluated and compared with the state of the art, obtaining satisfactory results in the implementation of embedded hardware.

Made available in DSpace on 2014-12-17T14:55:09Z (GMT). No. of bitstreams: 1 JoaoPFG_DISSERT.pdf: 1458535 bytes, checksum: fae41c0cfc4b0b09e65c1f69ec230a28 (MD5) Previous issue date: 2012-12-19 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior / A Quadrotor is an Unmanned Aerial Vehicle (UAV) equipped with four rotors distributed on a simple mechanical "X"form structure. The aim of this work is to build and stabilize a Quadrotor aircraft in the roll, pitch and yaw angles at a certain altitude. The stabilization control approach is based on a transformation in the input variables in order to perform a decoupled control. The proposed strategy is based on breaking the control problem into two hierarchical levels: A lower level, object of this work, maintains the desired altitude an angles of the vehicle while the higher level establishes appropriate references to the lower level, performing the desired movements. A hardware and software architecture was specially developed and implemented for an experimental prototype used to test and validate the proposed control approach / Um Quadrirrotor ? um Ve?culo A?reo N?o Tripulado (VANT) dotado de quatro rotores distribu?dos nas extremidades de uma estrutura mec?nica simples em forma de "X". O objetivo desse trabalho ? construir e estabilizar uma aeronave desse tipo, com alta capacidade de carga, em uma determinada altitude, sob ?ngulos de rolagem, guinada e arfagem pr?-definidos. A abordagem de controle de estabiliza??o baseia-se numa transforma??o das vari?veis de entrada do sistema a fim de realizar o controle de forma desacoplada. A estrat?gia proposta se baseia na divis?o do problema de controle em dois n?veis hier?rquicos: o n?vel inferior, objeto deste trabalho, mant?m os ?ngulos e a altitude do ve?culo em valores desejados, enquanto o n?vel superior estabelece refer?ncias adequadas para o n?vel inferior, de forma a executar os movimentos desejados Uma arquitetura de hardware e software foi especialmente desenvolvida e implementada para um prot?tipo experimental usado para testar e validar a abordagem de controle proposta

A computação visual é uma área do conhecimento que estuda o desenvolvimento de sistemas artificiais capazes de detectar e desenvolver a percepção do meio ambiente através de informações de imagem ou dados multidimensionais. A percepção visual e a manipulação são combinadas em sistemas robóticos através de duas etapas "olhar"e depois "movimentar-se", gerando um laço de controle de feedback visual. Neste contexto, existe um interesse crescimente no uso dessas técnicas em veículos aéreos não tripulados (VANTs), também conhecidos como drones. Essas técnicas são aplicadas para posicionar o drone em modo de vôo autônomo, ou para realizar a detecção de regiões para vigilância aérea ou pontos de interesse. Os sistemas de computação visual geralmente tomam três passos em sua operação, que são: aquisição de dados em forma numérica, processamento de dados e análise de dados. A etapa de aquisição de dados é geralmente realizada por câmeras e sensores de proximidade. Após a aquisição de dados, o computador embarcado realiza o processamento de dados executando algoritmos com técnicas de medição (variáveis, índice e coeficientes), detecção (padrões, objetos ou áreas) ou monitoramento (pessoas, veículos ou animais). Os dados processados são analisados e convertidos em comandos de decisão para o controle para o sistema robótico autônomo Visando realizar a integração dos sistemas de computação visual com as diferentes plataformas de VANTs, este trabalho propõe o desenvolvimento de um framework para controle de missão e guiamento de VANTs baseado em visão computacional. O framework é responsável por gerenciar, codificar, decodificar e interpretar comandos trocados entre as controladoras de voo e os algoritmos de computação visual. Como estudo de caso, foram desenvolvidos dois algoritmos destinados à aplicação em agricultura de precisão. O primeiro algoritmo realiza o cálculo de um coeficiente de reflectância visando a aplicação auto-regulada e eficiente de agroquímicos, e o segundo realiza a identificação das linhas de plantas para realizar o guiamento dos VANTs sobre a plantação. O desempenho do framework e dos algoritmos propostos foi avaliado e comparado com o estado da arte, obtendo resultados satisfatórios na implementação no hardware embarcado. / Cumputer Vision is an area of knowledge that studies the development of artificial systems capable of detecting and developing the perception of the environment through image information or multidimensional data. Nowadays, vision systems are widely integrated into robotic systems. Visual perception and manipulation are combined in two steps "look" and then "move", generating a visual feedback control loop. In this context, there is a growing interest in using computer vision techniques in unmanned aerial vehicles (UAVs), also known as drones. These techniques are applied to position the drone in autonomous flight mode, or to perform the detection of regions for aerial surveillance or points of interest. Computer vision systems generally take three steps to the operation, which are: data acquisition in numerical form, data processing and data analysis. The data acquisition step is usually performed by cameras or proximity sensors. After data acquisition, the embedded computer performs data processing by performing algorithms with measurement techniques (variables, index and coefficients), detection (patterns, objects or area) or monitoring (people, vehicles or animals). The resulting processed data is analyzed and then converted into decision commands that serve as control inputs for the autonomous robotic system In order to integrate the visual computing systems with the different UAVs platforms, this work proposes the development of a framework for mission control and guidance of UAVs based on computer vision. The framework is responsible for managing, encoding, decoding, and interpreting commands exchanged between flight controllers and visual computing algorithms. As a case study, two algorithms were developed to provide autonomy to UAVs intended for application in precision agriculture. The first algorithm performs the calculation of a reflectance coefficient used to perform the punctual, self-regulated and efficient application of agrochemicals. The second algorithm performs the identification of crop lines to perform the guidance of the UAVs on the plantation. The performance of the proposed framework and proposed algorithms was evaluated and compared with the state of the art, obtaining satisfactory results in the implementation of embedded hardware.

I compared forest structural parameters of treated and untreated plots on a forest restoration site on Galiano Island, British Columbia. The site was replanted with Douglas-fir (Pseudotsuga menziesii (mirb.) Franco) after being intensively logged in the 1970s and then thinned in the early 2000s. I used existing baseline data from 8 permanent plots (5 treated, 3 control) and compared it with forest assessment data collected in the field in the summer of 2017. Additionally, I used 16 temporary plots (8 treated, 8 control). I assessed vegetation percentage cover by plot, coarse woody debris by plot, tree diameter, species and status (n = 846), height (n = 48) and diameter growth (n = 271). I found that treated plots showed improved measures of structural diversity like diameter growth, crown ratios and plant diversity, but I was unable to relate the increased diameter growth to the restoration treatments. My findings suggest that to create a lasting impact, restoration thinning will have to be more frequent or create larger gaps. I then reviewed the current studies with unmanned aerial vehicles (UAV) in ecological restoration. I evaluated potential use of hobbyist UAVs for small organizations and not-for-profits and found that if applied correctly, UAVs can increase the amount of available data before, during and after restoration. Reproducible and reliable results require trained personnel and calibrated sensors. UAVs can increase access to remote areas and decrease disturbance of sensitive ecosystems. Regulations, limited flight time and processing time remain important restrictions on UAV use and hobbyist UAVs have a limit availability of sensors and flight performance. Finally, I used images taken from a hobbyist UAV to assess forest structure of the restoration site on Galiano Island and compared my results with the ground measurements. I found a canopy height model (CHM) from UAV images underestimated mean tree height values for the study site on average by 10.2 metres, while also severely underestimating mean stem densities. Using a 2 metre threshold, I delineated canopy gaps which accounted for 6 % of the canopy. UAV images and the resulting CHM represent a new visualization of the study site’s structure and can be a helpful tool in the communication of restoration outcomes to a wider audience. They are not, however, sufficient for monitoring or scientific applications. / Graduate

O advento dos veículos aéreos não tripulados (VANTs) representa uma quebra de paradigma no ramo aeronáutico. São revisados os conceitos de projeto envolvidos no desenvolvimento de VANTs e levantadas as suas potenciais aplicações. São também analisadas as características de sua operação e discutidas as questões regulatórias envolvidas na certificação e integração dos VANTs ao tráfego aéreo civil. Dentre as várias características singulares dos VANTs, enfatiza-se a sua necessidade de realizar vôos a grande altitude e com longa autonomia. Isso leva ao emprego de materiais mais leves e configurações com asas de grande alongamento, que provocam menor arrasto induzido. É feita uma revisão sobre materiais compósitos, que são materiais de uso crescente no ramo aeronáutico por sua leveza e resistência, e que por sua característica de anisotropia, são suscetíveis à otimização estrutural. Métodos e ferramentas de otimização estrutural de compósitos laminados ainda são pouco empregados, por suas características discretas e pelo grande número de parâmetros envolvidos. Um método eficiente e adequado à otimização de um problema desse tipo é o método dos algoritmos genéticos (AG). Assim foi desenvolvida uma sub-rotina de otimização baseada em algoritmos genéticos, usando a linguagem de programação Fortran. A sub-rotina desenvolvida trabalha em conjunto com um programa comercial de análise estrutural baseado no método dos elementos finitos, o Ansys. Foi também proposta uma configuração de asa de um VANT típico, de grande alongamento e fabricada com materiais compósitos reforçados com fibras (CRF). A asa proposta teve o número e a orientação das camadas do laminado otimizadas com o uso da sub-rotina desenvolvida, e resultados satisfatórios foram encontrados. Foram também analisados os efeitos da variação dos parâmetros dos operadores do AG, como probabilidades de mutação, cruzamento, tipo de escalonamento, entre outros, no desempenho do mesmo. Também foram feitas otimizações no mesmo modelo de asa proposto, empregando outros métodos disponíveis no próprio programa de análise estrutural. Os resultados das otimizações através desses métodos foram comparados com os resultados obtidos com a sub-rotina desenvolvida / The advent of the unmanned air vehicles (UAVs) represents a paradigm break in the aeronautical field. The project concepts involved in the development of UAVs are revised and its potential applications are rose. Also the issues related to UAVs operation are assessed and the regulatory questions involved in their certification and integration to the civil air space are argued. Amongst many singular characteristics of the UAV, its necessity of performing flights at high altitudes and with long endurance is emphasized. This leads to the employment of lighter materials and to configurations with high aspect ratio wings that cause minor induced drag. A revision on composite materials, which are of increasing use in the aeronautical field for their lightness and strength, and that are appropriate for being optimized due to their anisotropy characteristics, is made. Methods and tools of structural optimization of laminated composites are still seldom employed, due to their discrete nature and to the large number of parameters involved. An efficient and suitable method for the optimization of this kind of problem is the genetic algorithm (GA). Thus, an optimization sub-routine based on genetic algorithms was developed, using FORTRAN programming language. The developed sub-routine works in combination with Ansys, a structural analysis commercial program based on the finite elements method. A configuration of a typical UAV wing, made from composite reinforced plastics (CRP) was also proposed. The proposed wing had the number of plies and the orientations of its layers optimized using the developed sub-routine, and satisfactory results had been found. Also the effect of the variation of AG’s operator parameters in its performance, as mutation probabilities, crossover probabilities, fitness scaling, among others, have been assessed. The same wing model considered was also optimized using other build-in methods of the structural analysis program. The results of these optimizations have been compared with the results obtained with the developed sub-routine

O uso de veículos aéreos não tripulados (VANTs) tem se tornado cada vez mais comum, principalmente em aplicações de uso civil. No cenário militar, o uso de VANTs tem focado o cumprimento de missões específicas que podem ser divididas em duas grandes categorias: sensoriamento remoto e transporte de material de emprego militar. Este trabalho se concentra na categoria do sensoriamento remoto. O trabalho foca a definição de um modelo e uma arquitetura de referência para o desenvolvimento de sensores inteligentes orientados a missões específicas. O principal objetivo destas missões é a geração de mapas temáticos. Neste trabalho são investigados processos e mecanismos que possibilitem a geração desta categoria de mapas. Neste sentido, o conceito de MOSA (Mission Oriented Sensor Array) é proposto e modelado. Como estudos de caso dos conceitos apresentados são propostos dois sistemas de mapeamento automático de fontes sonoras, um para o caso civil e outro para o caso militar. Essas fontes podem ter origem no ruído gerado por grandes animais (inclusive humanos), por motores de combustão interna de veículos ou por atividade de artilharia (incluindo caçadores). Os MOSAs modelados para esta aplicação são baseados na integração de dados provenientes de um sensor de imageamento termal e uma rede de sensores acústicos em solo. A integração das informações de posicionamento providas pelos sensores utilizados, em uma base cartográfica única, é um dos aspectos importantes tratados neste trabalho. As principais contribuições do trabalho são a proposta de sistemas MOSA, incluindo conceitos, modelos, arquitetura e a implementação de referência representada pelo sistema de mapeamento automático de fontes sonoras. / The use of unmanned aerial vehicles (UAV) has become increasingly common, particularly in civilian applications. In the military scenario, the use of UAVs has focused the accomplishment of specific tasks in two broad categories: remote sensing and transport of military material. This work focuses the remote sensing category. It address the definition of a model and reference architecture for the development of smart sensors oriented to specific tasks. The main objective of these missions is to generate thematic maps. This work investigates processes and mechanisms that enable the automatic generation of thematic maps. In this sense, the concept of MOSA (Mission Oriented Sensor Array) is proposed and modeled. As case studies, we propose two automatic mapping systems for on-the-ground generated sound sources, one for the civilian case and one for the military case. These sounds may come from the noise generated by large animals (including humans), from internal combustion engine vehicles or from artillery activity (including hunters). The MOSAs modeled for this application integrate data from a thermal imaging sensor and an on-the-ground network of acoustic sensors. The fusion of position information, provided by the two sensors, into a single cartographic basis is one of the key aspects addressed in this work. The main contributions are the proposed MOSA systems, including concepts, models and architecture and the reference implementation comprised by the system for automatic mapping of sound sources.

The thesis covers a new navigation algorithm for UAV to fly through several given GPS coordinates without any human interference. The UAV first gets its current position from GPS receiver via Bluetooth connection with the navigator computer. With this GPS point, it draws an optimal trajectory to next destination. During the flight, the navigator computer issues the information about which direction to turn and how much to turn. This information will be used to steer the airplane servos. The algorithm is programmed in Java LeJOS. It uses built-in Java classes about GPS and Bluetooth. The main computer, where the navigation program runs, is a LEGO Mindstorms NXT and it is used a GPSlim240 from HOLUX as a GPS receiver.

En embarquant un appareil photo grand public de bonne qualité sur un drone ou un ULM, il est aujourd'hui possible de réaliser de manière 100% automatique des modélisations 3D visuellement « parfaites » ou en tout cas suffisante tant que ce modèle est utilisé comme support de communication. Par contre, les performances de ces méthodes sont encore mal maîtrisées dans le contexte où l'on veut utiliser le modèle comme outil de mesure. Les géométries d'acquisition linéaires, propices aux dérives de bandes photogrammétriques, en sont un exemple représentatif. Un phénomène de courbure de bande est alors constaté, nécessitant un travail de mesure terrain conséquent pour être corrigé. Ces imprécisions limitent l'intérêt de telles acquisitions, et sont un défi pour la communauté des photogrammètres, le marché naissant du drone, mais aussi de multiples industriel ayant un intérêt fort pour des systèmes capables d'assurer un suivi de mouvement 3D des sols avec une précision de quelques millimètres (par exemple suivi de digues, érosion agraire, surveillance de réseaux ferrés,...). Cette thèse est un projet de recherche issu d'un partenariat entre l'Institut National de l'Information Géographique et Forestière (IGN), et la Compagnie Nationale du Rhône (CNR), concessionnaire chargé d'entretenir et surveiller le réseau de digues de retenue du fleuve. Il s'agit - à partir de moyens aériens légers - d'arriver à un système d'auscultation plus rapide, plus économique et offrant une géométrie plus fine que les systèmes topométriques utilisés actuellement. Nous présentons tout d'abord une étude comparative des moyens aériens légers, montrant l'intérêt et les limitations des drones face aux ULM. Nous proposons des procédures d'acquisition adaptées aux géométries étudiées, permettant de limiter de manière opérationnelle les dérives. Dans un second temps, nous montrons que les phénomènes de courbure sont liés à des modèles d'orientation internes inadaptés, et nous proposons une procédure d'auto-calibration en trois étapes. Nos essais menés sur des chantiers linéaires montrent que le calcul des orientations est sensiblement amélioré (d'écarts métriques à des écarts centimétriques). Dans un troisième temps, nous proposons une optimisation de la compensation sur les points d'appuis, permettant de diminuer le nombre de mesures nécessaires. Enfin, nous présentons des méthodes pour contrôler l'incertitude des modèles 3D générés, et à travers une analyse diachronique, des exemples de suivi d'évolution permettant de suivre finement l'évolution de tels ouvrages / By embarking a good quality consumer grade camera on an Unmanned Aerial Vehicle (UAV) or on an ultralight system, it is possible to create – fully automatically – 3D models that are visually ''perfect'', or at least sufficient for visual or communication purposes. However, the accuracy of these methods is still a concern, especially if the resulting models are to be used as measuring tools. Linear acquisitions – also known as corridor mapping – are a striking example of the uncontrolled error propagation. A bowl effect is often observed, requiring multiple ground measurements to be overcome. These effects restrict the interest of such acquisitions. They are a challenge for the photogrammetrists community, the emerging civil UAV market, as well as many industrials interested in systems able to monitor ground deformations with an accuracy within a few millimeters (railway network surveillance, erosion of agricultural land, embankments monitoring,…).This thesis is an applied research project, led with a river concessionary which has the responsibility to maintain and keep under surveillance its containment dykes network. The purpose is to use light aerial systems as a mean for a monitoring system which is faster, cheaper, and more detailed than the topometric solutions used so far. First of all, we present a comparative study to understand, between UAVs and ultralight system, which tool is the most adapted for a specific need. We propose refined acquisition procedures, than can be used to limit operationally the drifts. In a second phase, we identify the internal camera model as the weak link, and propose a three steps auto-calibration procedure. Our tests led on linear acquisitions demonstrate that the external orientations accuracy are noticeably improved (from decimetric or even metric disparities, reduced to a few centimeters or less). In a third part, we propose an optimization of the adjustment on the ground control points, which is a way to limit the amount of ground measurements necessary. Finally, we present methods to control 3D models uncertainty, and through a diachronic analysis, examples of how such data can be used to monitor such structures through time

Arbetet gick ut på att utvärdera befintliga spårningsmetoder, utforma en procedur för spårning, ge förslag på lämplig utrustning och om tid fanns, testa konceptet. Arbetet skulle också dokumenteras så att krav, design och beslut kan förstås av någon som inte varit delaktig i arbetet. För att minska vibrationerna i UAV behöver rotorbladen spåras, det vill säga, rotorbladen ska befinna sig på samma höjd vid samma position i rotationen. Om spårningen inte är korrekt kan detta leda till oönskade vibrationer som kan vara farliga för UAV:er. I dagsläget används en metod som är väldigt tidskrävande. Metoden har heller inte speciellt god precision i mätningarna. För att komma fram med en ny spårningsmetod jämfördes olika metoder som används idag. Vilka krav som fanns på metoden, även kostnad och tid togs med vid utformandet. Spårningsmetoden som tagits fram innefattar en avståndsmätare som använder sig av ljus. Mätning sker på respektive rotorblad, därefter skickas data för behandling till en mikrokontroller. Resultatet blev ett nytt spårningskoncept med stor potential. Med mer utvecklingsarbete och testning skulle det kunna bli en egen produkt som kan till exepel säljas tillsammans med UAV eller helt separat. Examensarbetet har utförts hos företaget CybAero AB.