Centre dedocumentation
scientifique

Event-driven feature detection and tracking for visual SLAM / Ignacio Alzugaray (2022)  

Public Titre : Event-driven feature detection and tracking for visual SLAM Type de document : Thèse/HDR Auteurs : Ignacio Alzugaray, Auteur Editeur : Zurich : Eidgenossische Technische Hochschule ETH - Ecole Polytechnique Fédérale de Zurich EPFZ Année de publication : 2022 Note générale : bibliographie thesis submitted to attain the degree of Doctor of Sciences of ETH Zurich Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] caméra d'événement [Termes IGN] cartographie et localisation simultanées [Termes IGN] détection d'objet [Termes IGN] extraction de traits caractéristiques [Termes IGN] image floue [Termes IGN] reconnaissance de formes [Termes IGN] séquence d'images [Termes IGN] vision par ordinateur Index. décimale : THESE Thèses et HDR Résumé : (auteur) Traditional frame-based cameras have become the de facto sensor of choice for a multitude of applications employing Computer Vision due to their compactness, low cost, ubiquity, and ability to provide information-rich exteroceptive measurements. Despite their dominance in the field, these sensors exhibit limitations in common, real-world scenarios where detrimental effects, such as motion blur during high-speed motion or over-/underexposure in scenes with poor illumination, are prevalent. Challenging the dominance of traditional cameras, the recent emergence of bioinspired event cameras has opened up exciting research possibilities for robust perception due to their high-speed sensing, High-Dynamic-Range capabilities, and low power consumption. Despite their promising characteristics, event cameras present numerous challenges due to their unique output: a sparse and asynchronous stream of events, only capturing incremental perceptual changes at individual pixels. This radically different sensing modality renders most of the traditional Computer Vision algorithms incompatible without substantial prior adaptation, as they are initially devised for processing sequences of images captured at fixed frame-rate. Consequently, the bulk of existing event-based algorithms in the literature have opted to discretize the event stream into batches and process them sequentially, effectively reverting to frame-like representations in an attempt to mimic the processing of image sequences from traditional sensors. Such event-batching algorithms have demonstrably outperformed other alternative frame-based algorithms in scenarios where the quality of conventional intensity images is severely compromised, unveiling the inherent potential of these new sensors and popularizing them. To date, however, many newly designed event-based algorithms still rely on a contrived discretization of the event stream for its processing, suggesting that the full potential of event cameras is yet to be harnessed by processing their output more naturally. This dissertation departs from the mere adaptation of traditional frame-based approaches and advocates instead for the development of new algorithms integrally designed for event cameras to fully exploit their advantageous characteristics. In particular, the focus of this thesis lies on describing a series of novel strategies and algorithms that operate in a purely event-driven fashion, \ie processing each event as soon as it gets generated without any intermediate buffering of events into arbitrary batches and thus avoiding any additional latency in their processing. Such event-driven processes present additional challenges compared to their simpler event-batching counterparts, which, in turn, can largely be attributed to the requirement to produce reliable results at event-rate, entailing significant practical implications for their deployment in real-world applications. The body of this thesis addresses the design of event-driven algorithms for efficient and asynchronous feature detection and tracking with event cameras, covering alongside crucial elements on pattern recognition and data association for this emerging sensing modality. In particular, a significant portion of this thesis is devoted to the study of visual corners for event cameras, leading to the design of innovative event-driven approaches for their detection and tracking as corner-events. Moreover, the presented research also investigates the use of generic patch-based features and their event-driven tracking for the efficient retrieval of high-quality feature tracks. All the developed algorithms in this thesis serve as crucial stepping stones towards a completely event-driven, feature-based Simultaneous Localization And Mapping (SLAM) pipeline. This dissertation extends upon established concepts from state-of-the-art, event-driven methods and further explores the limits of the event-driven paradigm in realistic monocular setups. While the presented approaches solely rely on event-data, the gained insights are seminal to future investigations targeting the combination of event-based vision with other, complementary sensing modalities. The research conducted here paves the way towards a new family of event-driven algorithms that operate efficiently, robustly, and in a scalable manner, envisioning a potential paradigm shift in event-based Computer Vision. Note de contenu : 1- Introduction 2- Contribution 3- Conclusion and outlook Numéro de notice : 28699 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Thèse étrangère Note de thèse : PhD Thesis : Sciences : ETH Zurich : 2022 DOI : sans En ligne : https://www.research-collection.ethz.ch/handle/20.500.11850/541700 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100470

Benefiting from local rigidity in 3D point cloud processing / Zan Gojcic (2021)  

Public Titre : Benefiting from local rigidity in 3D point cloud processing Type de document : Thèse/HDR Auteurs : Zan Gojcic, Auteur Editeur : Zurich : Eidgenossische Technische Hochschule ETH - Ecole Polytechnique Fédérale de Zurich EPFZ Année de publication : 2021 Importance : 141 p. Format : 21 x 30 cm Note générale : bibliographie A thesis submitted to attain the degree of Doctor of Sciences of ETH Zurich Langues : Français (fre) Descripteur : [Vedettes matières IGN] Lasergrammétrie [Termes IGN] apprentissage profond [Termes IGN] capteur actif [Termes IGN] champ vectoriel [Termes IGN] déformation d'image [Termes IGN] données lidar [Termes IGN] effondrement de terrain [Termes IGN] enregistrement de données [Termes IGN] filtrage du bruit [Termes IGN] flux [Termes IGN] image 3D [Termes IGN] navigation autonome [Termes IGN] orientation du capteur [Termes IGN] segmentation [Termes IGN] semis de points [Termes IGN] téléphone intelligent [Termes IGN] traitement de nuage de points [Termes IGN] voxel Index. décimale : THESE Thèses et HDR Résumé : (auteur) Incorporating 3D understanding and spatial reasoning into (intelligent) algorithms is crucial for solving several tasks in fields such as engineering geodesy, risk assessment, and autonomous driving. Humans are capable of reasoning about 3D spatial relations even from a single 2D image. However, making the priors that we rely on explicit and integrating them into computer programs is very challenging. Operating directly on 3D input data, such as 3D point clouds, alleviates the need to lift 2D data into a 3D representation within the task-specific algorithm and hence reduces the complexity of the problem. The 3D point clouds are not only a better-suited input data representation, but they are also becoming increasingly easier to acquire. Indeed, nowadays, LiDAR sensors are even integrated into consumer devices such as mobile phones. However, these sensors often have a limited field of view, and hence multiple acquisitions are required to cover the whole area of interest. Between these acquisitions, the sensor has to be moved and pointed in a different direction. Moreover, the world that surrounds us is also dynamic and might change as well. Reasoning about the motion of both the sensor and the environment, based on point clouds acquired in two-time steps, is therfore an integral part of point cloud processing. This thesis focuses on incorporating rigidity priors into novel deep learning based approaches for dynamic 3D perception from point cloud data. Specifically, the tasks of point cloud registration, deformation analysis, and scene flow estimation are studied. At first, these tasks are incorporated into a common framework where the main difference is in the level of rigidity assumptions that are imposed on the motion of the scene or the acquisition sensor. Then, the tasks specific priors are proposed and incorporated into novel deep learning architectures. While the global rigidity can be assumed in point cloud registration, the motion patterns in deformation analysis and scene flow estimation are more complex. Therefore, the global rigidity prior has to be relaxed to local or instancelevel rigidity, respectively. Rigidity priors not only add structure to the aforementioned tasks, which prevents physically implausible estimates and improves the generalization of the algorithms, but in some cases also reduce the supervision requirements. The proposed approaches were quantitatively and qualitatively evaluated on several datasets, and they yield favorable performance compared to the state-of-the-art. Numéro de notice : 28660 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Thèse étrangère Note de thèse : PhD : Sciences : ETH Zurich : 2021 DOI : sans En ligne : https://www.research-collection.ethz.ch/handle/20.500.11850/523368 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=99817

Drivers and implications of dominant and rare tree species in global forests / Iris Hordijk (2021)  

Public Titre : Drivers and implications of dominant and rare tree species in global forests Type de document : Thèse/HDR Auteurs : Iris Hordijk, Auteur Editeur : Zurich : Eidgenossische Technische Hochschule ETH - Ecole Polytechnique Fédérale de Zurich EPFZ Année de publication : 2021 Note générale : bibliographie Thesis submitted to attain the degree of Doctor of Sciences of ETH Zurich Langues : Anglais (eng) Descripteur : [Termes IGN] âge du peuplement forestier [Termes IGN] biodiversité végétale [Termes IGN] biome [Termes IGN] écosystème forestier [Termes IGN] espèce végétale [Termes IGN] extinction (biologie) [Termes IGN] forêt boréale [Termes IGN] forêt tempérée [Termes IGN] forêt tropicale [Termes IGN] habitat (nature) [Termes IGN] peuplement forestier [Termes IGN] plante menacée [Termes IGN] productivité biologique [Termes IGN] richesse floristique [Termes IGN] service écosystémique [Vedettes matières IGN] Inventaire forestier Index. décimale : THESE Thèses et HDR Résumé : (auteur) Forests provide crucial ecosystem functions and services for the earth system and humanity. Due to ongoing deforestation and forest degradation, these ecosystems are increasingly fragmented and disturbed, significantly changing tree species composition within the forest. Most plant communities are comprised of a few dominant species that are numerically abundant, and many rare species, each of which exist at low abundance. Given their differences in abundance and characteristics, the loss of dominant and rare species have distinct impacts on ecosystem functioning. Decreasing abundances of dominant species typically have larger immediate impacts on overall ecosystem processes (e.g. productivity), while a decline in abundance of rare species can lead to the loss of key functions and overall multifunctionality, and can ultimately drive species extinctions. Despite their unique role in the ecosystem, it remains unclear what drives species to become dominant or rare, how threatened locally dominant and rare species are, and what the effect of species abundance on ecosystem function is across large environmental gradients in our global forests. The aim of this thesis is to explore the drivers of tree species abundance, and evaluate the effect of relative tree species abundance on forest productivity. Specifically, the three chapters of this thesis aimed to identify at a global scale the 1) patterns, drivers, and threats to dominant and rare tree species, 2) differences in trait values and trait diversity that differentiate the functional contributions of dominant and rare tree species, and 3) how evenness (the relative species abundances in the community) mediates the relationship between tree species richness and forest productivity in forests. In this thesis I answered these research questions by analyzing a global dataset of forest composition, which enabled me to describe broad-scale ecological patterns and to test general ecological laws. Note de contenu : General introduction 1- Patterns, drivers and threats to dominant and rare tree species worldwide 2- Trait diversity of dominant and rare tree species in global forests: a habitat filtering perspective 3- Evenness mediates the global relationship between forest productivity and richness Synthesis Numéro de notice : 28690 Affiliation des auteurs : non IGN Thématique : BIODIVERSITE/FORET Nature : Thèse étrangère Note de thèse : PhD Thesis : Sciences : ETH Zurich 2021 DOI : sans En ligne : https://www.research-collection.ethz.ch/handle/20.500.11850/520710 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100131

From point clouds to high-fidelity models - advanced methods for image-based 3D reconstruction / Audrey Richard (2021)  

Public Titre : From point clouds to high-fidelity models - advanced methods for image-based 3D reconstruction Type de document : Thèse/HDR Auteurs : Audrey Richard, Auteur Editeur : Zurich : Eidgenossische Technische Hochschule ETH - Ecole Polytechnique Fédérale de Zurich EPFZ Année de publication : 2021 Note générale : bibliographie A thesis submitted to attain the degree of Doctor of Sciences of ETH Zurich Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] apprentissage profond [Termes IGN] chaîne de traitement [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] compréhension de l'image [Termes IGN] démonstration de faisabilité [Termes IGN] discrétisation spatiale [Termes IGN] jeu de données localisées [Termes IGN] modèle 3D de l'espace urbain [Termes IGN] modélisation sémantique [Termes IGN] optimisation (mathématiques) [Termes IGN] Pays-Bas [Termes IGN] reconstruction 3D [Termes IGN] reconstruction d'objet [Termes IGN] Rhénanie du Nord-Wesphalie (Allemagne) [Termes IGN] texturage [Termes IGN] Zurich (Suisse) Résumé : (auteur) Capturing automatically a virtual 3D model of an object or a scene from a collection of images is a useful capability with a wide range of applications, including virtual/augmented reality, heritage preservation, consumer digital entertainment, autonomous robotics, navigation, industrial vision or metrology, and many more. Since the early days of photogrammetry and computer vision, it has been a topic of intensive research but has eluded a general solution for it. 3D modeling requires more than reconstructing a cloud of 3D points from images; it requires a high-fidelity representation whose form is often dependent on individual objects. This thesis guides you in the journey of image-based 3D reconstruction through several advanced methods that aims to push its boundaries, from precise and complete geometry to detailed appearance, using both theory with elegant mathematics and more recent breakthroughs in deep learning. To evaluate these methods, thorough experiments are conducted at scene level (and large-scale) where efficiency is of key importance, and at object level where accuracy, completeness and photorealism can be better appreciated. To show the individual potential of each of these methods, as well as the possible wide coverage in terms of applications, different scenarios are considered and serve as a proof-of-concept. Thereby, the journey starts with large-scale city modeling using aerial photography from the cities of Zürich (Switzerland), Enschede (Netherlands) and Dortmund (Germany), followed by single object completion using the synthetic dataset ShapeNet, that includes objects like cars, benches or planes that can be found in every city, to finish with the embellishment of these digital models via high-resolution texture mapping using a multi-view 3D dataset of real and synthetic objects, like for example statues and fountains that also dress the landscape of cities. Combining them together into an incremental pipeline dedicated to a specific application would require further tailoring but is quite possible. Numéro de notice : 17650 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Thèse étrangère Note de thèse : PhD : Sciences : ETH Zurich : 2021 En ligne : http://dx.doi.org/10.3929/ethz-b-000461735 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=97892

Collaborative visual-inertial state and scene estimation / Marco Karrer (2020)  

Public Titre : Collaborative visual-inertial state and scene estimation Type de document : Thèse/HDR Auteurs : Marco Karrer, Auteur ; Margarita Chli, Directeur de thèse Editeur : Zurich : Eidgenossische Technische Hochschule ETH - Ecole Polytechnique Fédérale de Zurich EPFZ Année de publication : 2020 Importance : 151 p. Format : 21 x 30 cm Note générale : bibliographie A thesis submitted to attain the degree of Doctor of Sciences of ETH Zurich in Mechanical Engineering Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Acquisition d'image(s) et de donnée(s) [Termes IGN] cartographie et localisation simultanées [Termes IGN] centrale inertielle [Termes IGN] compensation par faisceaux [Termes IGN] estimation de pose [Termes IGN] image captée par drone [Termes IGN] reconstruction d'objet [Termes IGN] robotique [Termes IGN] système multi-agents [Termes IGN] vision par ordinateur Index. décimale : THESE Thèses et HDR Résumé : (auteur) The capability of a robot to create a map of its workspace on the fly, while constantly updating it and continuously estimating its motion in it, constitutes one of the central research problems in mobile robotics and is referred to as Simultaneous Localization And Mapping (SLAM) in the literature. Relying solely on the sensor-suite onboard the robot, SLAM is a core building block in enabling the navigational autonomy necessary to facilitate the general use of mobile robots and has been the subject of booming research interest spanning over three decades. With the largest body of related literature addressing the challenge of single-agent SLAM, it is only very recently, with the relative maturity of this field that approaches tackling collaborative SLAM with multiple agents have started appearing. The potential of collaborative multi-agent SLAM is great; not only promising to boost the efficiency of robotic missions by splitting the task at hand to more agents but also to improve the overall robustness and accuracy by boosting the amount of data that each agent’s estimation process has access to. While SLAM can be performed using a variety of different sensors, this thesis is focused on the fusion of visual and inertial cues, as one of the most common combinations of sensing modalities in robotics today. The information richness captured by cameras, along with the high-frequency and metric information provided by Inertial Measurement Units (IMUs) in combination with the low weight and power consumption offered by a visual-inertial sensor suite render this setup ideal for a wide variety of applications and robotic platforms, in particular to resource-constrained platforms such as Unmanned Aerial Vehicles (UAVs). The majority of the state-of-the-art visual-inertial estimators are designed as odometry algorithms, providing only estimates consistent within a limited time-horizon. This lack in global consistency of estimates, however, poses a major hurdle in an effective fusion of data from multiple agents and the practi- cal definition of a common reference frame, which is imperative before collaborative effort can be coordinated. In this spirit, this thesis investigates the potential of global optimization, based on a central access point (server) as a first approach, demonstrating global consistency using only monocular-inertial data. Fusing data from multiple agents, not only consistency can be maintained, but also the accuracy is shown to improve at times, revealing the great potential of collaborative SLAM. Aiming at improving the computational efficiency, in a second approach a more efficient system architecture is employed, allowing a more suitable distribution of the computational load amongst the agents and the server. Furthermore, the architecture implements a two-way communication enabling a tighter collaboration between the agents as they become capable of re-using information captured by other agents through communication with the server, enabling improvements of their onboard pose tracking online, during the mission. In addition to general collaborative SLAM without specific assumptions on the agents’ relative pose configuration, we investigate the potential of a configuration with two agents, carrying one camera each with overlapping fields of view, essentially forming a virtual stereo camera. With the ability of each robotic agent to move independently, the potential to control the stereo baseline according to the scene depth is very promising, for example at high altitudes where all scene points are far away and, therefore, only provide weak constraints on the metric scale in a standard single-agent system. To this end, an approach to estimate the time-varying stereo transformation formed between two agents is proposed, by fusing the egomotion estimates of the individual agents along with the image measurements extracted from the view-overlap in a tightly coupled fashion. Taking this virtual stereo camera idea a step further, a novel collaboration framework is presented, utilizing the view-overlap along with relative distance measurements across the two agents (e.g. obtained via Ultra-Wide Band (UWB) modules), in order to successfully perform state estimation at high altitudes where state-of-the-art single-agent methods fail. In the interest of low-latency pose estimation, each agent holds its own estimate of the map, while consistency between the agents is achieved using a novel consensus-based sliding window bundle adjustment. Despite that in this work, experiments are shown in a two-agent setup, the proposed distributed bundle adjustment scheme holds great potential for scaling up to larger problems with multiple agents, due to the asynchronicity of the proposed estimation process and the high level of parallelism it permits. The majority of the developed approaches in this thesis rely on sparse feature maps in order to allow for efficient and timely pose estimation, however, this translates to reduced awareness of the spatial structure of a robot’s workspace, which can be insufficient for tasks requiring careful scene interaction and manipulation of objects. Equipping a typical visual-inertial sensor suite with an RGB-D camera, an add-on framework is presented that enables the efficient fusion of naturally noisy depth information into an accurate, local, dense map of the scene, providing sufficient information for an agent to plan contact with a surface. With the focus on collaborative SLAM using visual-inertial data, the approaches and systems presented in this thesis contribute towards achieving collaborative Visual-Inertial SLAM (VI-SLAM) deployable in challenging real-world scenarios, where the participating agents’ experiences get fused and processed at a central access point. On the other side, it is shown that taking advantage of specific configurations can push the collaboration amongst the agents towards achieving greater general robustness and accuracy of scene and egomotion estimates in scenarios, where state-of-the-art single-agent systems are otherwise unsuccessful, paving the way towards intelligent robot collaboration. Note de contenu : Introduction 1- Real-time dense surface reconstruction for aerial manipulation 2- Towards globally consistent visual-inertial collaborative SLAM 3- CVI-SLAM – collaborative visual-inertial SLAM 4- Collaborative 6DoF relative pose estimation for two UAVs with overlapping fields of view 5- Distributed variable-baseline stereo SLAM from two UAVs Numéro de notice : 28318 Affiliation des auteurs : non IGN Thématique : IMAGERIE/INFORMATIQUE Nature : Thèse étrangère Note de thèse : PhD Thesis : Mechanical Engineering : ETH Zurich : 2020 DOI : sans En ligne : https://www.research-collection.ethz.ch/handle/20.500.11850/465334 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=98251

Enhancing knowledge, skills, and spatial reasoning through location-based mobile learning / Christian Sailer (2020)  

Permalink

Modelling historical accessibility and its effects in space / Raphaël Fuhrer (2020)  

Permalink

Point cloud registration and mitigation of refraction effects for geomonitoring using long-range terrestrial laser scanning / Ephraim Friedli (2020)  

Permalink

Realistic modeling of power transmission lines with geographic information systems / Joram Schito (2020)  

Permalink

Convolutional neural network for traffic signal inference based on GPS traces / Yann Méneroux (2018) 

Permalink

Spatial big data and machine learning in GIScience, Workshop at GIScience 2018, Melbourne, Australia, 28 August 2018 / Martin Raubal (2018)    

Permalink

Monitoring rock glaciers by combining photogrammetric and GNSS-based methods / Fabian Neyer (2017)  

Permalink

Advanced modeling and algorithms for high-precision GNSS analysis / Kan Wang (2016)  

Permalink

Autonomous navigation in complex nonplanar environments based on laser ranging / Philipp Andreas Krüsi (2016)  

Permalink

Co-location of geodetic observation techniques in space / Benjamin Männel (2016)  

Permalink