No description available.

La reconstruction de bâtiments à partir de photographies aériennes et d’autres données spatiales urbaines multi-sources est une tâche qui utilise une multitude de méthodes automatisées et semi-automatisées allant des processus ponctuels au traitement classique des images et au balayage laser. Dans cette thèse, un système de relaxation itératif est développé sur la base de l'examen du contexte local de chaque bord en fonction de multiples sources d'entrée spatiales (masques optiques, d'élévation, d'ombre et de feuillage ainsi que d'autres données prétraitées, décrites au chapitre 6). Toutes ces données multisource et multirésolution sont fusionnées de manière à extraire les segments de ligne probables ou les arêtes correspondant aux limites des bâtiments. Deux nouveaux sous-systèmes ont également été développés dans cette thèse. Ils ont été conçus dans le but de fournir des informations supplémentaires, plus fiables, sur les contours des bâtiments dans une future version du système de relaxation proposé. La première est une méthode de réseau de neurones à convolution profonde (CNN) pour la détection de frontières de construction. Le réseau est notamment basé sur le modèle SRCNN (Dong C. L., 2015) de super-résolution à la pointe de la technologie. Il accepte des photographies aériennes illustrant des données de zones urbaines densément peuplées ainsi que leurs cartes d'altitude numériques (DEM) correspondantes. La formation utilise trois variantes de cet ensemble de données urbaines et vise à détecter les contours des bâtiments grâce à une nouvelle cartographie hétéroassociative super-résolue. Une autre innovation de cette approche est la conception d'une couche de perte personnalisée modifiée appelée Top-N. Dans cette variante, l'erreur quadratique moyenne (MSE) entre l'image de sortie reconstruite et l'image de vérité de sol (GT) fournie des contours de bâtiment est calculée sur les 2N pixels de l'image avec les valeurs les plus élevées. En supposant que la plupart des N pixels de contour de l’image GT figurent également dans les 2N pixels supérieurs de la reconstruction, cette modification équilibre les deux catégories de pixels et améliore le comportement de généralisation du modèle CNN. Les expériences ont montré que la fonction de coût Top-N offre des gains de performance par rapport à une MSE standard. Une amélioration supplémentaire de la capacité de généralisation du réseau est obtenue en utilisant le décrochage. Le deuxième sous-système est un réseau de convolution profonde à super-résolution, qui effectue un mappage associatif à entrée améliorée entre les images d'entrée à basse résolution et à haute résolution. Ce réseau a été formé aux données d’altitude à basse résolution et aux photographies urbaines optiques à haute résolution correspondantes. Une telle différence de résolution entre les images optiques / satellites optiques et les données d'élévation est souvent le cas dans les applications du monde réel. / Building reconstruction from aerial photographs and other multi-source urban spatial data is a task endeavored using a plethora of automated and semi-automated methods ranging from point processes, classic image processing and laser scanning. In this thesis, an iterative relaxation system is developed based on the examination of the local context of each edge according to multiple spatial input sources (optical, elevation, shadow & foliage masks as well as other pre-processed data as elaborated in Chapter 6). All these multisource and multiresolution data are fused so that probable line segments or edges are extracted that correspond to prominent building boundaries.Two novel sub-systems have also been developed in this thesis. They were designed with the purpose to provide additional, more reliable, information regarding building contours in a future version of the proposed relaxation system. The first is a deep convolutional neural network (CNN) method for the detection of building borders. In particular, the network is based on the state of the art super-resolution model SRCNN (Dong C. L., 2015). It accepts aerial photographs depicting densely populated urban area data as well as their corresponding digital elevation maps (DEM). Training is performed using three variations of this urban data set and aims at detecting building contours through a novel super-resolved heteroassociative mapping. Another innovation of this approach is the design of a modified custom loss layer named Top-N. In this variation, the mean square error (MSE) between the reconstructed output image and the provided ground truth (GT) image of building contours is computed on the 2N image pixels with highest values . Assuming that most of the N contour pixels of the GT image are also in the top 2N pixels of the re-construction, this modification balances the two pixel categories and improves the generalization behavior of the CNN model. It is shown in the experiments, that the Top-N cost function offers performance gains in comparison to standard MSE. Further improvement in generalization ability of the network is achieved by using dropout.The second sub-system is a super-resolution deep convolutional network, which performs an enhanced-input associative mapping between input low-resolution and high-resolution images. This network has been trained with low-resolution elevation data and the corresponding high-resolution optical urban photographs. Such a resolution discrepancy between optical aerial/satellite images and elevation data is often the case in real world applications. More specifically, low-resolution elevation data augmented by high-resolution optical aerial photographs are used with the aim of augmenting the resolution of the elevation data. This is a unique super-resolution problem where it was found that many of -the proposed general-image SR propositions do not perform as well. The network aptly named building super resolution CNN (BSRCNN) is trained using patches extracted from the aforementioned data. Results show that in comparison with a classic bicubic upscale of the elevation data the proposed implementation offers important improvement as attested by a modified PSNR and SSIM metric. In comparison, other proposed general-image SR methods performed poorer than a standard bicubic up-scaler.Finally, the relaxation system fuses together all these multisource data sources comprising of pre-processed optical data, elevation data, foliage masks, shadow masks and other pre-processed data in an attempt to assign confidence values to each pixel belonging to a building contour. Confidence is augmented or decremented iteratively until the MSE error fails below a specified threshold or a maximum number of iterations have been executed. The confidence matrix can then be used to extract the true building contours via thresholding.

In Geographic Information System (GIS) applications, aster data constitutes one of the major data types. The displaying of the raster data has an important part in GIS applications. Digital Terrain Elevation Data (DTED) is one of the raster data types, which is used as the main data source in this thesis. The DTED data is displayed on the screen as digital images as a pixel value, which is represented in gray scale, corresponding to an elevation (texel). To draw the images, the texel values are mostly interpolated in order to perform zoom-in and/or zoom-out operations on the concerned area. We implement and compare four types of interpolation methods, nearest neighbor, bilinear interpolation, and two new proposed interpolation methods (1) 4-texel weighted average and (2) 8-texel weighted average. The real-time graphical display, with zoom-in/zoom-out capabilities, has also been implemented by buffering DTED data in memory and using a C++ clas that manages graphical operations (zoom-in, zoom-out, and 2D, 3D isplay) by using Windows GDI+ and OpenGL graphic ibraries resulting in 30-40 framesper-second for one grid of DTED Level 0 data.

The Egrigoz, Koyunoba, and Alacam plutons are located in the Northern Menderes Massif of western Turkey between the Simav normal fault to the south and the Izmir-Ankara-Erzincan suture to the north. Although much attention has focused on their geochemical and geochronological history, their relationship to each other and other major structures in the region is still debated. Some geologic maps show the Egrigoz and Koyunoba pluton bounded to the west by the low-angle Simav detachment fault. In contrast, other regional maps show no offsets between the plutons and surrounding metamorphic rocks. Yet other studies indicate thrust faults may be present near the Egrigoz pluton, between Menderes metamorphic rocks and a meta-rhyolite unit. To gain a better understanding of the history of the Egrigoz, Koyunoba, and Alacam plutons, ArcGIS digital elevation data from the region, geochronological data, geochemical analyses, and cathodoluminescence (CL) images were acquired to search for effects of micro- to macro-scales of deformation. Numerous ~E-W trending extension lineations that parallel the Simav graben and cut the plutons were observed in relief images. These lineations, likely due to large-scale ~N-S extension, continue across plutons inferring that extension continued after the exhumation of these rocks. The Simav graben and its associated high-angle fault are evident in the elevation data, but no other significant detachment-related basins or structures are shown, including the low-angle Simav detachment. U-Pb zircon ages, ranging from 29.9±3.9 Ma to 14.6±2.6 Ma, suggest the plutons crystallized over a ~15 m.y. time frame. Samples from the plutons are peraluminous S-type granite to granodiorites. The plutons were emplaced in a post-collisional volcanic-arc setting and range from magnesian to ferroan with increasing silica contents. Geochemical analyses show little difference between the three plutons, consistent with the rocks arising from a similar source. To document microstructures that might help explain these heterogeneities, CL images were obtained. CL images document a complicated tectonic history including magma mixing, multiple episodes of brittle deformation, and fluid alteration. The CL images constitute evidence of a complex multi-stage tectonic history for the region that includes water-mediated brittle deformation. / text

In this thesis, an unmanned air vehicle (UAV) is used to develop a prototype base test platform for flight testing of new control algorithms and avionics for advanced UAV system development applications. A control system that holds the UAV at a fixed altitude above the ground is designed and flight tested. Only the longitudinal motion of the UAV is considered during the controller design, hence its lateral motions are controlled manually by a remote control unit from the ground. UAV&amp / #8217 / s altitude with respect to the mean sea level and position are obtained by an onboard global positioning system (GPS) and this information is transmitted to the ground computer via radio frequency (RF) communication modules. The altitude of the UAV above the ground is calculated by using the digital terrain elevation data (DTED). A controller is designed and its gains are tuned to maintain this flight altitude at a desired value by using the mathematical model developed to represent the longitudinal dynamics of the UAV. Input signals generated by the controller for elevator deflections are transmitted back to the UAV via RF communication modules to drive onboard servomotors to generate desired elevator deflections. All controller computations and RF communications are handled by a MATLAB&reg / based platform on a ground computer. UAV flight tests are carried out at two different autopilot modes / namely, mean sea level (MSL) altitude hold mode and above ground level (AGL) altitude hold mode. The developed platform worked properly during flight tests and proved to be reliable in almost every condition. Moreover, the designed controller system is demonstrated to be effective and it fulfills the requirements.

Airborne angle-only geolocalization is the localization of objects on ground level from airborne vehicles (AV) using bearing measurements, namely azimuth and elevation. This thesis aims to introduce elevation data of the terrain to the airborne angle-only geolocalization problem and to demonstrate that it could be applicable for localization of jammers. Jammers are often used for deliberate interference with malicious intent which could interfere with the positioning system of a vehicle. It is important to locate the jammers to either avoid them or to remove them.    Three localization methods, i.e. the nonlinear least squares (NLS), the extended Kalman filter (EKF) and the unscented Kalman filter (UKF), are implemented and tested on simulated data. The methods are also compared to the theoretical lower bound, the Cramér-Rao Lower Bound (CRLB), to see if there is an efficient estimator. The simulated data are different scenarios where the number of AVs, the relative flight path of the AVs and the knowledge of the terrain can differ. Using the knowledge of the terrain elevation, the methods give more consistent localization than without it. Without elevation data, the localization relies on good geometry of the problem, i.e. the relative flight path of the AVs, while the geometry is not as critical when elevation data is available. However, the elevation data does not always improve the localization for certain geometries.    There is no method that is clearly better than the others when elevation data is used. The methods’ performances are very similar and they all converge to the CRLB but that could also be an advantage. This makes the usage of elevation data not restricted to a certain method and it leaves more up to the implementer which method they prefer.

This thesis is about using airborne laser scanning data for making maps for Orienteering. In this thesis were used altimetry data DMR 4G and DMR 5G. These data are provided by ČUZK. The control measuring was made in choosen area to verify the accuracy. In this thesis is described procedure of data collection and their treatment. The results of this thesis are two illustrations of the orienteering maps.