In this dissertation, we have focused on conflicts that occur due to disagreeing motions in multi-modal localization algorithms. In spite of the recent achievements in robust localization by means of multi-sensor fusion, these algorithms are not applicable to all environments. This is primarily attributed to the following fundamental assumptions: (i) the environment is predominantly stationary, (ii) only ego-motion of the sensor platform exists, and (iii) multiple sensors are always in agreement with each other regarding the observed motion. Recently, studies have shown how to relax the static environment assumption using outlier rejection techniques and dynamic object segmentation. Additionally, to handle non ego-motion, approaches that extend the localization algorithm to multi-body tracking have been studied. However, there has been no attention given to the conditions where multiple sensors contradict each other with regard to the motions observed. Vision based localization has become an attractive approach for both indoor and outdoor applications due to the large information bandwidth provided by images and reduced cost of the cameras used. In order to improve the robustness and overcome the limitations of vision, an Inertial Measurement Unit (IMU) may be used. Even though visual-inertial localization has better accuracy and improved robustness due to the complementary nature of camera and IMU sensor, they are affected by disagreements in motion observations. We term such dynamic situations as environments with motion conflictbecause these are caused when multiple different but self- consistent motions are observed by different sensors. Tightly coupled visual inertial fusion approaches that disregard such challenging situations exhibit drift that can lead to catastrophic errors. We have provided a probabilistic model for motion conflict. Additionally, a novel algorithm to detect and resolve motion conflicts is also presented. Our method to detect motion conflicts is based on per-frame positional estimate discrepancy and per- landmark reprojection errors. Motion conflicts were resolved by eliminating inconsistent IMU and landmark measurements. Finally, a Motion Conflict aware Visual Inertial Odometry (MC- VIO) algorithm that combined both detection and resolution of motion conflict was implemented. Both quantitative and qualitative evaluation of MC-VIO on visually and inertially challenging datasets were obtained. Experimental results indicated that MC-VIO algorithm reduced the absolute trajectory error by 70% and the relative pose error by 34% in scenes with motion conflict, in comparison to the reference VIO algorithm. Motion conflict detection and resolution enables the application of visual inertial localization algorithms to real dynamic environments. This paves the way for articulate object tracking in robotics. It may also find numerous applications in active long term augmented reality.

This thesis proposes a motion sensing system for wheelchairs with the main objective of determining tips, falls and risky situations. The system relies on measurements from an Inertial Measurement Unit, (IMU), consisting of a 3-axis accelerometer and a 2-axis gyroscope as the source of information. The IMU was embedded in a portable device, the "Motion Logger", which collects motion data in a Secure Digital memory card after running a real time preprocessing algorithm. The algorithm was designed to reduce energy consumption and memory usage. Actual signal analysis and attitude estimation is carried out offline. The motion sensing system was developed for determining wheelchair-related falls as part of a major research effort carried out at the research center of the James A Haley VA Hospital Subject Safety Center, Tampa, Florida. The focus of the study concentrated on achieving a thorough understanding of the demographics, nature, consequences and the creation of prediction models for fall events. The main goal of the embedded system was to successfully estimate the motion variables relevant to the occurrence of falls, tips and similar risky situations. Currently, off-line smoothing techniques based on Kalman filter concepts allow for optimal estimation of angles in the longitudinal direction, roll, and in the lateral direction, pitch. Results from both predefined experiments with known outcomes and data collected from actual wheelchair users during pilot and final deployment stages are presented and discussed.

Cette thèse aborde le problème de la calibration d’un ensemble de capteurscomposé d’une centrale inertielle, d’un magnétomètre et d’une caméra, avecpour objectif leur intégration sur un système très compact : un mini-drone.Cette étude expose tout d’abord les contraintes imposées par l’application surle choix des capteurs et les solutions envisagées notamment pour résoudre leproblème de la synchronisation des mesures. Après avoir étudié les techniquesde calibration existantes, une méthode permettant la calibration de l’ensembledes capteurs (accéléromètre, gyromètre, magnétomètre et caméra) est présentée.La solution proposée permet également d’estimer les changements de repèresentre les différents capteurs. Elle a la particularité de ne nécessiter l’emploid’aucun matériel particulier. D’autre part, l’intégration de ces capteurs dans unsystème aussi compact soulève de nouvelles difficultés. Dans ces conditions, leschamps magnétiques créés par les actionneurs du drone perturbent les mesuresdu magnétomètre se trouvant à proximité. Une nouvelle méthode est proposéeafin d’estimer et de compenser dynamiquement ces perturbations magnétiquesen fonction de l’état des actionneurs du drone. Enfin, deux applications dusystème comprenant une centrale inertielle et une caméra sont présentées :la construction de mosaïques d’images géo-référencées et la stabilisation devidéos. Ces deux applications exploitent les mesures des capteurs inertiels afind’effectuer un traitement en temps réel pour un coût calculatoire très faible. / This thesis deal with the issue of the calibration of a group of sensor composedof an inertial unit, a magnetometer and a camera. It aims at integratingthem into a very compact system : a mini-drone. First of all, this study outlinesthe constraints imposed by the application on the choice of the sensors andthe solutions considered to solve the measures synchronization issue. Afterstudying existing calibration techniques, a method for the calibration of allthe sensors (accelerometer, gyroscope, magnetometer and camera) is presented.The proposed solution allows to estimate the frame transformation between thedifferent sensors. It has the advantage of not requiring the use of any specialequipment. Furthermore, the integration of these sensors into a compact systemraises new difficulties. Under these conditions, the magnetic fields created bythe drone actuators disrupt magnetometer measurements. A new method isproposed to estimate and compensate for these magnetic disturbances. Thecompensation is dynamically adapted based on the state of the drone actuators.Finally, two applications of the system including an inertial unit and a cameraare presented : the construction of geo-referenced images mosaic and videostabilization. Both applications use measurements of inertial sensors and precisecalibration to perform a real-time processing for a very low computational cost.

L'estimation de l'attitude (ou orientation) est un problème récurrent de nombreuses applications allant de la robotique aérienne ou sous-marine en passant par des applications médicales (surveillance de patients, réhabilitation), mais aussi jeux vidéo, etc. L'objectif de cette thèse est d'évaluer l'apport des approches ensemblistes dans le cadre de l'estimation de l'attitude à partir de données issues de triaxes accéléromètres (A), magnétomètres (M) et gyromètres (G). Dans un premier temps, on s'intéresse aux mouvements "quasi-statiques" et l'estimation de l'attitude est réalisée à partir de mesures AM. On aborde ensuite le cas des mouvements dynamiques, en considérant l'ensemble des mesures AGM. Le problème du choix de la paramétrisation de l'attitude a été abordé et on a comparé les résultats obtenus et le temps calcul pour des modélisations avec les angles de Cardan et le quaternion unitaire. Les algorithmes développés ont été validés en simulation et avec des données réelles. Les résultats ont été comparés avec ceux fournis par des algorithmes de l'état de l'art, par exemple SIVIA. La deuxième partie du manuscrit est consacrée à au diagnostic des capteurs de la centrale inertielle avec des approches ensemblistes. Les algorithmes développés dans la première partie du travail sont adaptés afin de pouvoir détecter et localiser un défaut dans l'ensemble des capteurs considérés.

Attitude and Heading angles are crucial parameters for navigation. Conventional navigation methods mostly uses IMU and GPS devices to calculate these angles. MEMS technology offers small sized, low cost IMU sensors with moderate performance. However, GPS cannot be used in underwater. Therefore, different aiding sensors are used in underwater vehicles in order to increase the accuracy. As the accuracy of devices increases, the cost of these devices also increases. In this thesis, rather than using GPS and high quality IMU sensors, low cost MEMS IMU sensor is used together with a magnetometer and a pressure sensor as aiding sensors. Considering the IMU error model and motion dynamics, two systems are designed and simulated using real data. The results seem to be satisfactory and using pressure sensor as an aiding sensor improves the attitude angles estimation.

Today's smartphones are equipped with a variety of different sensors such as GPS receivers, accelerometers, gyroscopes and magnetometers, making smartphones viable tools in many applications. The computational capacity of smartphones allows for software applications running advanced signal processing algorithms. Thus, attaching a smartphone inside a car makes it possible to estimate kinematics of the vehicle by fusing information from the different sensors inside the smartphone. Fusing information from different sources for improving estimation quality is a well-known problem and there exist a lot of methods and algorithms in this area. This thesis approaches the sensor fusion problem of estimating kinematics of cars using smartphones for the purpose of analysing driving performance. Different varieties of the coordinated turn model for describing the vehicle dynamics are investigated. Also, different measurement models are evaluated where bias errors of the sensors are taken into consideration. Pre-filtering and construction of pseudo-measurements are also considered which allow for use of state space models with a lower dimension. / Dagens smartphones är utrustade med en rad olika typer av sensorer såsom GPS mottagare, accelerometrar, gyroskop och magnetometrar vilket medför ett stort användningsområde. Beräkningskapaciteten hos smartphones gör det möjligt för mjukvaruapplikationer att använda sig av avancerade algoritmer för signalbehandling. Det är därför möjligt att placera en smartphone inuti en bil och skatta bilens kinematik genom att kombinera informationen från de olika sensorerna. Att fusionera information från olika källor för att erhålla bättre skattningar är ett välkänt område där det finns många metoder och algoritmer utvecklade. Detta examensarbete behandlar sensorfusionsproblemet att skatta bilars kinematik med hjälp av smartphones för syftet att kunna analysera körprestanda. Olika varianter av en coordinated turn modell för att beskriva bilens dynamik undersöks. Dessutom testas olika modeller för sensorerna där hänsyn till exempelvis biasfel tas. Förbehandling av data och pseudomätningar testas också vilket gör det möjligt att använda tillståndsmodeller med låg dimension.

This thesis presents the design of an inertial measurement data acquisition system intended for use in sailboats. The variables of interest are 3-axis acceleration, 3-axis rotation, GPS position/velocity, magnetic compass bearing and wind speed/direction. The design focus is on low-cost micro-electromechanical systems (MEMS) based technology and demonstrating the validity of these technologies in a scientific application. A prototype is constructed and submitted to a series of tests to demonstrate functionality and soundness of the design. These tests range from bench tests to full scale application. Contributions of this thesis include the novel application of inertial measurement unit (IMU) technology to a sailboat racing application, the integration of all instrumentation, creative ruggedised packaging and emphasising the use of low-cost commercial off-the-shelf (COTS) technology.

The loss of a hand or arm is a devastating life event that results in many months of healing and challenging rehabilitation. Technology has allowed the development of an electronic replacement for a lost limb but similar advancements in therapy have not occurred. The situation is made more challenging because people with amputations often do not live near specialized rehabilitation centres. As a result, delays in therapy can worsen common complications like nerve pain and joint stiffness. For children born without a limb, poor compliance with the use of their prosthesis leads to delays in therapy and may affect their development. In many parts of the world, amputation rehabilitation does not exist. Fortunately, we live in an age where advances in technology and engineering can help solve these problems. Virtual reality creates a simulated world or environment through computer animation much like what is seen in modern video games. An experienced team of rehabilitation doctors, therapists, engineers and computer scientists are required to realize a system such as this. A person with an amputation will be taught to control objects in the virtual world by wearing a modified electronic prosthesis. Using computers, it will be possible to analyze his or her movements within the virtual world and improve the wearer's skills. The goals of this system include making the system portable and internet compatible so that people living in remote areas can also receive therapy. The novel approach of using virtual reality to rehabilitate people with upper limb amputations will help them return to normal activities by providing modern and appropriate rehabilitation, reducing medical complications, improving motivation (via gaming modules), advancing health care technology and reducing health care costs. The use of virtual reality technology in the field of amputee rehabilitation is in its earliest stages of development world wide. A virtual environment (VE) will facilitate the early rehabilitation of a patient before they are clinically ready to be fitted with an actual prosthesis. In order to create a successful virtual reality rehabilitation system such as this, an accurate method of tracking the arm in real-time is necessary. A linear displacement sensor and a microelectromechanical system (MEMS) inertial measurement unit (IMU) were used to create a device for capturing the motion of a user's movement with the intent that the data provided by the device be used along with a VE as a virtual rehabilitation tool for new upper extremity amputation patients. This thesis focuses on the design and testing of this motion capture device in order to determine the suitability of current commercially available sensing components as used in this system. Success will be defined by the delivery of accurate position and orientation data from the device so that that data can be used in a virtual environment. Test results show that with current MEMS sensors, the error introduced by double integrating acceleration data is too significant to make an IMU an acceptable choice for position tracking. However, the device designed here has proven to be an excellent cable emulator, and would be well suited if used as an orientation tracker.

In order to detect movements and vibrations on different appliance, especially robots and self-going devices are used different sensors. One of the most used movement sensors are accelerometers. They are three different types of accelerometers one-axes, two-axes and three-axes. They can be analogue or digital. Husqvarna AB uses an accelerometer to detect movements of their self-going lawnmower Automower. Since it begins to turn up others sensors on the market and the sensor which today uses in Husqvarna’s lawnmower don’t fulfils all functions there is a need to find other solutions. What sensors can be found on a market and which new has come? How can they be used best and which difference is between them? How can they use to detect movements in x, y and z direction and how can they read different movements to detect collision, inclination and lift? Sensor analysis will be done to find out how they can be used in this matter and an big analysis will be done to find out how a sensor witch already is in a lawnmower can be use on the better ways? / För att kunna detektera rörelse och vibrationer på olika anordningar och framförallt robotar och självgående maskiner används olika sensorer. En av de mest använda rörelsesensorer är accelerometer. Det finns en uppsjö av olika utföringar som kan vara med en axel, två axlar eller tre axlar. Accelerometrar finns som digitala eller analoga. Husqvarna AB använder en sådan sensor för att detektera rörelse på sin självgående gräsklippare Automower. Eftersom det börjar dyka upp andra sensorer på marknaden och den sensor som används idag uppfyller inte alla funktioner finns det behov av att lägga tid på att se vad kan åstadkommas med en sådan sensor. Vilka sensorer kan hittas på marknaden och vilka nya har dykt upp? Hur kan de användas på bästa sätt och vad skiljer de åt? Hur kan de användas för att detektera rörelse i x, y, z-led och hur kan de läsa av olika rörelse för att detektera krock, lutning och lyft? Det skall göras analys för att se hur olika sensorer reagerar och hur kan de användas på bästa sätt. Det skall utforskas hur den sensor som redan sitter kan användas på ett bättre sätt.

Positioning and navigation systems play an important role in our daily life, but now most of positioning systems were confined in outdoor environments, most of which were used on transportation. Therefore, the goal of this thesis is to develop a Pedestrian Dead Reckoning System (PDRS), which can not only be used to solve a problem of GPS out-of-lock, but also be used in the field of indoor positioning. In dangerous environments, such as the scene of a fire, when the rescue personnel have an accident on himself or discover a wounded who need to be salvaged, if the rescue personnel who has configured the PDRS, then the other rescue personnel can assist them immediately. In the part of hardware system, we used embedded system to be the primary part of the entire system, the embedded system has the characters of low power consumption and portability. Therefore, we chose the TI OMAP35x EVM platform to be our primary system of PDRS. In order to get the information of pedestrian, we also need the Inertial Measurement Unit (IMU) and Compass to provide the information of acceleration and heading for PDRS. To achieve the function of remote monitoring, we used wireless transmission module to send data of sensors to OMAP35x EVM. Finally, the most important function that we must accomplish in this thesis is to use OMAP35x EVM to build a real-time PDRS. In the part of software system, we use Linux OS and Qt SDK to build the software system of PDRS in this thesis. In the part of algorithm, we use step detection, step length estimation and dead reckoning method to construct the algorithm of PDRS in this thesis.