Le raffinement de nuage de points et la reconstruction de surface sont deux problèmes fondamentaux dans le traitement de la géométrie. La plupart des méthodes existantes ont été ciblées sur les données de capteur de distance et se sont avérées être mal adaptées aux données multi-vues. Dans cette thèse, deux nouvelles méthodes sont proposées respectivement pour les deux problèmes avec une attention particulière aux données multi-vues. La première méthode permet de lisser les nuages de points provenant de la reconstruction multi-vue sans endommager les données. Le problème est formulé comme une optimisation non-linéaire sous contrainte et ensuite résolu par une série de problèmes d’optimisation sans contrainte au moyen d’une méthode de barrière. La seconde méthode effectue une triangulation du nuage de points d’entrée pour générer un maillage en utilisant une stratégie de l’avancement du front pilotée par un critère de l’empilement compact de sphères. L’algorithme est simple et permet de produire efficacement des maillages de haute qualité. Les expérimentations sur des données synthétiques et du monde réel démontrent la robustesse et l’efficacité des méthodes proposées. Notre méthodes sont adaptées aux applications qui nécessitent des informations de position précises et cohérentes telles que la photogrammétrie et le suivi des objets en vision par ordinateur / Point cloud refinement and surface reconstruction are two fundamental problems in geometry processing. Most of the existing methods have been targeted at range sensor data and turned out be ill-adapted to multi-view data. In this thesis, two novel methods are proposed respectively for the two problems with special attention to multi-view data. The first method smooths point clouds originating from multi-view reconstruction without impairing the data. The problem is formulated as a nonlinear constrained optimization and addressed as a series of unconstrained optimization problems by means of a barrier method. The second method triangulates point clouds into meshes using an advancing front strategy directed by a sphere packing criterion. The method is algorithmically simple and can produce high-quality meshes efficiently. The experiments on synthetic and real-world data have been conducted as well, which demonstrates the robustness and the efficiency of the methods. The developed methods are suitable for applications which require accurate and consistent position information such photogrammetry and tracking in computer vision

The term ‘Industry 4.0’ leads to many new possibilities like smart factory which is the amalgamation of manufacturing systems in a network to perform tasks more efficiently. It is becoming more and more important for the companies to develop smart factories and integrate the devices within such a facility to meet the demands of the evolving market. The next generation production systems are designed to share the data within the network, plan, and predict the solution for the future problems. One such technology under smart factory is 3D laser scanning resulting in point cloud of the production unit. The traditional way of documenting a layout is usually with the help of 2D computer aided designs which are susceptible to measurement errors and changes that are not updated regularly. With the help of point clouds, an as-is representation of the factories can be recorded which can be easily updated with changes in the real world. With advancements in virtual manufacturing, the need for visualization of the factories is increasing drastically. 3D Laser Scanning is one of the better ways of meeting this need, among many other applications. The focus of the thesis had been to create a method document for 3D laser scanning of factories and to discuss the future possibilities of it. The research approach used in this thesis was conducting observational study, interviews and testing of the method. One such future possibility is autonomous scanning and how it would be beneficial for a company like Scania which is developing smart factories. Based on the study carried out during the thesis, a document presenting the method developed is included in the report. The report also points out the applications and benefits of point cloud over traditional layout planning methods.

Submitted by Silvana Teresinha Dornelles Studzinski (sstudzinski) on 2015-10-27T14:34:35Z No. of bitstreams: 1 Fabrício Müller da Silva_.pdf: 105910980 bytes, checksum: 4ce66a9d5fff9a2b2a97835c54dac355 (MD5) / Made available in DSpace on 2015-10-27T14:34:35Z (GMT). No. of bitstreams: 1 Fabrício Müller da Silva_.pdf: 105910980 bytes, checksum: 4ce66a9d5fff9a2b2a97835c54dac355 (MD5) Previous issue date: 2015-08-31 / Nenhuma / Este trabalho apresenta um algoritmo para simplificação de nuvens de pontos baseado na inclinação local da superfície amostrada pelo conjunto de pontos de entrada. O objetivo é transformar a nuvem de pontos original no menor conjunto possível, mantendo as características e a topologia da superfície original. O algoritmo proposto reamostra de forma adaptativa o conjunto de entrada, removendo pontos redundantes para manter um determinado nível de qualidade definido pelo usuário no conjunto final. O processo consiste em um particionamento recursivo do conjunto de entrada através da Análise de Componentes Principais (PCA). No algoritmo, PCA é aplicada para definir as partições sucessivas, para obter uma aproximação linear (por planos) em cada partição e para avaliar a qualidade de cada aproximação. Por fim, o algoritmo faz uma escolha simples de quais pontos serão mantidos para representar a aproximação linear de cada partição. Estes pontos formarão o conjunto de dados final após o processo de simplificação. Para avaliação dos resultados foi aplicada uma métrica de distância entre malhas de polígonos, baseada na distância de Hausdorff, comparando a superfície reconstruída com a nuvem de pontos original e aquela reconstruída com a nuvem filtrada. Os resultados obtidos com o algoritmo conseguiram uma taxa de até 95% de compactação do conjunto de dados de entrada, diminuindo o tempo total de execução do processo de reconstrução, mantendo as características e a topologia do modelo original. A qualidade da superfície reconstruída com a nuvem filtrada também é atestada pela métrica de comparação. / This paper presents a simple and efficient algorithm for point cloud simplification based on the local inclination of the surface sampled by the input set. The objective is to transform the original point cloud in a small as possible one, keeping the features and topology of the original surface. The proposed algorithm performs an adaptive resampling of the input set, removing unnecessary points to maintain a level of quality defined by the user in the final dataset. The process consists of a recursive partitioning in the input set using Principal Component Analysis (PCA). PCA is applied for defining the successive partitions, for obtaining the linear approximations (planes) for each partition, and for evaluating the quality of those approximations. Finally, the algorithm makes a simple choice of the points to represent the linear approximation of each partition. These points are the final dataset of the simplification process. For result evaluation, a distance metric between polygon meshes, based on Hausdorff distance, was defined, comparing the reconstructed surface using the original point clouds and the reconstructed surface usingthe filtered ones. The algorithm achieved compression rates up to 95% of the input dataset,while reducing the total execution time of reconstruction process, keeping the features and the topology of the original model. The quality of the reconstructed surface using the filtered point cloud is also attested by the distance metric.

Traditional forest management relies on a small field sample and interpretation of aerial photography that not only are costly to execute but also yield inaccurate estimates of the entire forest in question. Airborne light detection and ranging (LiDAR) is a remote sensing technology that records point clouds representing the 3D structure of a forest canopy and the terrain underneath. We present a method for segmenting individual trees from the LiDAR point clouds without making prior assumptions about tree crown shapes and sizes. We then present a method that vertically stratifies the point cloud to an overstory and multiple understory tree canopy layers. Using the stratification method, we modeled the occlusion of higher canopy layers with respect to point density. We also present a distributed computing approach that enables processing the massive data of an arbitrarily large forest. Lastly, we investigated using deep learning for coniferous/deciduous classification of point cloud segments representing individual tree crowns. We applied the developed methods to the University of Kentucky Robinson Forest, a natural, majorly deciduous, closed-canopy forest. 90% of overstory and 47% of understory trees were detected with false positive rates of 14% and 2% respectively. Vertical stratification improved the detection rate of understory trees to 67% at the cost of increasing their false positive rate to 12%. According to our occlusion model, a point density of about 170 pt/m² is needed to segment understory trees located in the third layer as accurately as overstory trees. Using our distributed processing method, we segmented about two million trees within a 7400-ha forest in 2.5 hours using 192 processing cores, showing a speedup of ~170. Our deep learning experiments showed high classification accuracies (~82% coniferous and ~90% deciduous) without the need to manually assemble the features. In conclusion, the methods developed are steps forward to remote, accurate quantification of large natural forests at the individual tree level.

La maquette numérique de bâtiments historiques (Heritage-BIM) est devenue un enjeu majeur dans la modélisation. Le modèle HBIM ainsi obtenu comprend une description paramétrique et sémantique des éléments constitutifs du patrimoine. La thèse présente des méthodes de constructions HBIM à partir de la documentation historique, de nuages de points, de maillage de surfaces et de géométrie solide. Un concept de mesh-to-HBIM est proposé à l'aide de la programmation visuelle, qui permet de transformer les « familles » paramétriques et les structures géométriques en modèles paramétriques et sémantiques HBIM. La modélisation paramétrique HBIM consiste à créer manuellement des Familles Revit paramétriques et une reconstruction de bâtiment semi-automatisée par l'application de scripts Dynamo. Le processus de modélisation sémantique HBIM transforme directement des géométries segmentées de maillages ou de solides vers l'environnement BIM. Les éléments segmentés et individualisés peuvent être stockés et gérés dans cet environnement avec des compléments d'informations d'association entre éléments. / Heritage Building Information Modelling (HBIM) is a major issue in heritage documentation and conservation. The obtained HBIM model provides a parametric and semantic description of the heritage elements. This thesis presents methods for the generation of HBIM models from point clouds (obtained by photogrammetry or laser scanning), surface mesh and solid geometry. A concept of solid/mesh-to-HBIM is proposed using Autodesk Dynamo visual programming, which transfers the parametric “Family” and geometric structures to parametric and semantic HBIM models. The parametric HBIM modelling process involves conventional manual parametric “Family” creation and semi-automated building reconstruction by Dynamo. The semantic HBIM modelling process directly transfers the segmented solid geometry and closed mesh-to-BIM environment. The segmented elements can be stored and managed in the BIM environment with attached attributes information and relationships established among the elements.

Cette recherche vise à la modélisation 3D depuis un système de vision fisheye embarqué, utilisée pour une application GNSS dans le cadre du projet Predit CAPLOC. La propagation des signaux satellitaires en milieu urbain est soumise à des réflexions sur les structures, altérant la précision et la disponibilité de la localisation. L’ambition du projet est (1) de définir un système de vision omnidirectionnelle capable de fournir des informations sur la structure 3D urbaine et (2) de montrer qu’elles permettent d’améliorer la localisation.Le mémoire expose les choix en (1) calibrage automatique, (2) mise en correspondance entre images, (3) reconstruction 3D ; chaque algorithme est évalué sur images de synthèse et réelles. De plus, il décrit une manière de corriger les réflexions des signaux GNSS depuis un nuage de points 3D pour améliorer le positionnement. En adaptant le meilleur de l’état de l’art du domaine, deux systèmes sont proposés et expérimentés. Le premier est un système stéréoscopique à deux caméras fisheye orientées vers le ciel. Le second en est l’adaptation à une unique caméra.Le calibrage est assuré à travers deux étapes : l’algorithme des 9 points adapté au modèle « équisolide » couplé à un RANSAC, suivi d’un affinement par optimisation Levenberg-Marquardt. L’effort a été porté sur la manière d’appliquer la méthode pour des performances optimales et reproductibles. C’est un point crucial pour un système à une seule caméra car la pose doit être estimée à chaque nouvelle image.Les correspondances stéréo sont obtenues pour tout pixel par programmation dynamique utilisant un graphe 3D. Elles sont assurées le long des courbes épipolaires conjuguées projetées de manière adaptée sur chaque image. Une particularité est que les distorsions ne sont pas rectifiées afin de ne pas altérer le contenu visuel ni diminuer la précision. Dans le cas binoculaire il est possible d’estimer les coordonnées à l’échelle. En monoculaire, l’ajout d’un odomètre permet d’y arriver. Les nuages successifs peuvent être calés pour former un nuage global en SfM.L’application finale consiste dans l’utilisation du nuage 3D pour améliorer la localisation GNSS. Il est possible d’estimer l’erreur de pseudodistance d’un signal après multiples réflexions et d’en tenir compte pour une position plus précise. Les surfaces réfléchissantes sont modélisées grâce à une extraction de plans et de l’empreinte des bâtiments. La méthode est évaluée sur des paires d’images fixes géo-référencées par un récepteur bas-coût et un récepteur GPS RTK (vérité terrain). Les résultats montrent une amélioration de la localisation en milieu urbain. / This research deals with 3D modelling from an embedded fisheye vision system, used for a GNSS application as part of CAPLOC project. Satellite signal propagation in urban area implies reflections on structures, impairing localisation’s accuracy and availability. The project purpose is (1) to define an omnidirectional vision system able to provide information on urban 3D structure and (2) to demonstrate that it allows to improve localisation.This thesis addresses problems of (1) self-calibration, (2) matching between images, (3) 3D reconstruction ; each algorithm is assessed on computer-generated and real images. Moreover, it describes a way to correct GNSS signals reflections from a 3D point cloud to improve positioning. We propose and evaluate two systems based on state-of-the-art methods. First one is a stereoscopic system made of two sky facing fisheye cameras. Second one is the adaptation of the former to a single camera.Calibration is handled by a two-steps process: the 9-point algorithm fitted to “equisolid” model coupled with a RANSAC, followed by a Levenberg-Marquardt optimisation refinement. We focused on the way to apply the method for optimal and repeatable performances. It is a crucial point for a system composed of only one camera because the pose must be estimated for every new image.Stereo matches are obtained for every pixel by dynamic programming using a 3D graph. Matching is done along conjugated epipolar curves projected in a suitable manner on each image. A distinctive feature is that distortions are not rectified in order to neither degrade visual content nor to decrease accuracy. In the binocular case it is possible to estimate full-scale coordinates.In the monocular case, we do it by adding odometer information. Local clouds can be wedged in SfM to form a global cloud.The end application is the usage of the 3D cloud to improve GNSS localisation. It is possible to estimate and consider a signal pseudodistance error after multiple reflections in order to increase positioning accuracy. Reflecting surfaces are modelled thanks to plane and buildings trace fitting. The method is evaluated on fixed image pairs, georeferenced by a low-cost receiver and a GPS RTK receiver (ground truth). Study results show the localisation improvement ability in urban environment.

This thesis examines an automated method to detect surface deformations on automotive panels for the purpose of quality control along a manufacturing assembly line. Automation in the automotive manufacturing industry is becoming more prominent, but quality control is still largely performed by human workers. Quality control is important in the context of automotive body panels as deformations can occur along the assembly line such as inadequate handling of parts or tools around a vehicle during assembly, rack storage, and shipping from subcontractors. These defects are currently identified and marked, before panels are either rectified or discarded. This work attempts to develop an automated system to detect deformations to alleviate the dependence on human workers in quality control and improve performance by increasing speed and accuracy. Some techniques make use of an ideal CAD model behaving as a master work, and panels scanned on the assembly line are compared to this model to determine the location of deformations. This thesis presents a solution for detecting deformations of various scales without a master work. It also focuses on automated analysis requiring minimal intuitive operator-set parameters and provides the ability to classify the deformations as dings, which are deformations that protrude from the surface, or dents, which are depressions into the surface. A complete automated deformation detection system is proposed, comprised of a feature extraction module, segmentation module, and classification module, which outputs the locations of deformations when provided with the 3D mesh of an automotive panel. Two feature extraction techniques are proposed. The first is a general feature extraction technique for 3D meshes using octrees for multi-resolution analysis and evaluates the amount of surface variation to locate deformations. The second is specifically designed for the purpose of deformation detection, and analyzes multi-resolution cross-sections of a 3D mesh to locate deformations based on their estimated size. The performance of the proposed automated deformation detection system, and all of its sub-modules, is tested on a set of meshes which represent differing characteristics of deformations in surface panels, including deformations of different scales. Noisy, low resolution meshes are captured from a 3D acquisition, while artificial meshes are generated to simulate ideal acquisition conditions. The proposed system shows accurate results in both ideal situations as well as non-ideal situations under the condition of noise and complex surface curvature by extracting only the deformations of interest and accurately classifying them as dings or dents.

This thesis examines an automated method to detect surface deformations on automotive panels for the purpose of quality control along a manufacturing assembly line. Automation in the automotive manufacturing industry is becoming more prominent, but quality control is still largely performed by human workers. Quality control is important in the context of automotive body panels as deformations can occur along the assembly line such as inadequate handling of parts or tools around a vehicle during assembly, rack storage, and shipping from subcontractors. These defects are currently identified and marked, before panels are either rectified or discarded. This work attempts to develop an automated system to detect deformations to alleviate the dependence on human workers in quality control and improve performance by increasing speed and accuracy. Some techniques make use of an ideal CAD model behaving as a master work, and panels scanned on the assembly line are compared to this model to determine the location of deformations. This thesis presents a solution for detecting deformations of various scales without a master work. It also focuses on automated analysis requiring minimal intuitive operator-set parameters and provides the ability to classify the deformations as dings, which are deformations that protrude from the surface, or dents, which are depressions into the surface. A complete automated deformation detection system is proposed, comprised of a feature extraction module, segmentation module, and classification module, which outputs the locations of deformations when provided with the 3D mesh of an automotive panel. Two feature extraction techniques are proposed. The first is a general feature extraction technique for 3D meshes using octrees for multi-resolution analysis and evaluates the amount of surface variation to locate deformations. The second is specifically designed for the purpose of deformation detection, and analyzes multi-resolution cross-sections of a 3D mesh to locate deformations based on their estimated size. The performance of the proposed automated deformation detection system, and all of its sub-modules, is tested on a set of meshes which represent differing characteristics of deformations in surface panels, including deformations of different scales. Noisy, low resolution meshes are captured from a 3D acquisition, while artificial meshes are generated to simulate ideal acquisition conditions. The proposed system shows accurate results in both ideal situations as well as non-ideal situations under the condition of noise and complex surface curvature by extracting only the deformations of interest and accurately classifying them as dings or dents.

Šiame darbe tiriami algoritminiai sprendimai mobiliam robotui, leidžiantys aptikti ieškomą objektą bei įvertinti jo poziciją ir orientaciją erdvėje. Atlikus šios srities technologijų analizę surasta įvairių realizacijai tinkamų metodų, tačiau bendro jų efektyvumo palyginimo trūko. Siekiant užpildyti šią spragą realizuota programinė ir techninė įranga, kuria atliktas labiausiai roboto sistemoms tinkamų metodų vertinimas. Algoritmų analizė susideda iš algoritmų tikslumo ir jų veikimo spartos vertinimo panaudojant tam paprastus bei efektyvius metodus. Darbe analizuojamas objektų orientacijos nustatymas iš Kinect kameros gylio duomenų pasitelkiant ICP algoritmą. Atliktas dviejų gylio sistemų spartos ir tikslumo tyrimas parodė, jog Kinect kamera spartos atžvilgiu yra efektyvesnis bei 2-5 kartus tikslesnis sprendimas nei įprastinė stereo kamerų sistema. Objektų aptikimo algoritmų efektyvumo eksperimentuose nustatytas maksimalus aptikimo tikslumas apie 90% bei pasiekta maksimali 15 kadrų/s veikimo sparta analizuojant standartinius VGA 640x480 raiškos vaizdus. Atliktas objektų pozicijos ir orientacijos nustatymo ICP metodo efektyvumo tyrimas parodė, jog vidutinė absoliutinė pozicijos ir orientacijos nustatymo paklaida yra atitinkamai apie 3.4cm bei apie 30 laipsnių, o veikimo sparta apie 2 kadrai/s. Tolesnis optimizavimas arba duomenų kiekio minimizavimas yra būtinas norint pasiekti geresnius veikimo rezultatus mobilioje ribotų resursų roboto sistemoje. Darbe taip pat buvo sėkmingai... [toliau žr. visą tekstą] / This work presents a performance analysis of the state-of-the-art computer vision algorithms for object detection and pose estimation. Initial field study showed that many algorithms for the given problem exist but still their combined comparison was lacking. In order to fill in the existing gap a software and hardware solution was created and the comparison of the most suitable methods for a robot system were done. The analysis consists of detector accuracy and runtime performance evaluation using simple and robust techniques. Object pose estimation via ICP algorithm and stereo vision Kinect depth sensor method was used in this work. A conducted two different stereo system analysis showed that Kinect achieves best runtime performance and its accuracy is 2-5 times more superior than a regular stereo setup. Object detection experiments showcased a maximum object detection accuracy of nearly 90% and speed of 15 fps for standard size VGA 640x480 resolution images. Accomplished object position and orientation estimation experiment using ICP method showed, that average absolute position and orientation detection error is respectively 3.4cm and 30 degrees while the runtime speed – 2 fps. Further optimization and data size minimization is necessary to achieve better efficiency on a resource limited mobile robot platform. The robot hardware system was also successfully implemented and tested in this work for object position and orientation detection.

This thesis examines an automated method to detect surface deformations on automotive panels for the purpose of quality control along a manufacturing assembly line. Automation in the automotive manufacturing industry is becoming more prominent, but quality control is still largely performed by human workers. Quality control is important in the context of automotive body panels as deformations can occur along the assembly line such as inadequate handling of parts or tools around a vehicle during assembly, rack storage, and shipping from subcontractors. These defects are currently identified and marked, before panels are either rectified or discarded. This work attempts to develop an automated system to detect deformations to alleviate the dependence on human workers in quality control and improve performance by increasing speed and accuracy. Some techniques make use of an ideal CAD model behaving as a master work, and panels scanned on the assembly line are compared to this model to determine the location of deformations. This thesis presents a solution for detecting deformations of various scales without a master work. It also focuses on automated analysis requiring minimal intuitive operator-set parameters and provides the ability to classify the deformations as dings, which are deformations that protrude from the surface, or dents, which are depressions into the surface. A complete automated deformation detection system is proposed, comprised of a feature extraction module, segmentation module, and classification module, which outputs the locations of deformations when provided with the 3D mesh of an automotive panel. Two feature extraction techniques are proposed. The first is a general feature extraction technique for 3D meshes using octrees for multi-resolution analysis and evaluates the amount of surface variation to locate deformations. The second is specifically designed for the purpose of deformation detection, and analyzes multi-resolution cross-sections of a 3D mesh to locate deformations based on their estimated size. The performance of the proposed automated deformation detection system, and all of its sub-modules, is tested on a set of meshes which represent differing characteristics of deformations in surface panels, including deformations of different scales. Noisy, low resolution meshes are captured from a 3D acquisition, while artificial meshes are generated to simulate ideal acquisition conditions. The proposed system shows accurate results in both ideal situations as well as non-ideal situations under the condition of noise and complex surface curvature by extracting only the deformations of interest and accurately classifying them as dings or dents.