L’objectif de la thèse est de développer un nouveau système de localisation mobile composé de trois caméras 3D de type Kinect et d’une caméra additionnelle de type Fish Eye. La solution algorithmique est basée sur l’odométrie visuelle et permet de calculer la trajectoire du mobile en temps réel à partir des données fournies par les caméras 3D. L’originalité de la méthodologie réside dans l’exploitation d’orthoimages créées à partir des nuages de points acquis en temps réel par les trois caméras. L’étude des différences entre les orthoimages successives acquises par le système mobile permet d’en déduire ses positions successives et d’en calculer sa trajectoire. / The aim of the thesis is to develop a new kind of localization system, composed of three 3D cameras such as Kinect and an additional Fisheye camera. The localization algorithm is based on Visual Odometry principles in order to calculate the trajectory of the mobile platform in real time from the data provided by the 3D cameras.The originality of the processing method lies within the exploitation of orthoimages processed from the point clouds that are acquired in real time by the three cameras. The positions and trajectory of the mobile platform can be derived from the study of the differences between successive orthoimages.

GPS/INS (Global Positioning System / Inertial Navigation System) brings possibility of relaxing the demand for aerial triangulation in stereo model construction and rectification. In this thesis a differential rectification algorithm is programmed with Matlab software for aerial frame camera images. This program is tested using exterior orientation parameters obtained by GPS/INS and images are ortho-rectified. Ground Control Points (GCP) are measured in the orthorectified images and compared with other rectification methods according to RMSE and mean error. Besides, direct georeferencing accuracy is investigated by using GPS/INS data. Therefore, stereo models and ortho-images are constructed by using exterior orientation parameters obtained with both aerial triangulation and GPS/INS integration. GCPs are measured in both stereo models and ortho-images, compared with respect to their RMSE and mean error. In order to determine Digital Elevation Model (DEM) effect in ortho-rectification, different DEM data are used and the results are compared.

There are a number of studies that deal with detection of road network from image Remote Sensing data. However, little work has been done on algorithms for horizontal road signs detection from Remote Sensing data. On the other hand there are many papers that deal with horizontal road signs detection in the field of Computer vision. Theoretical part of master thesis sumarizes the methods that are used to detect objects from both, Remote Sensing and Computer vision data. In the practical part an algorithm of automatic Zebra-crossing detection based on Aerial Images and vector road layer is designed. Zebra-crossings are detected using matching of image with set of predefined patterns. Obtained set of potential objects is then filtred using geometric and relational criteria. At the end of this thesis, the proposed algorithm is validated and the results are discussed with literature. Key words: Zebra-crossing, Horizontal road signs, Image matching, Aerial images

In this thesis, sensitivity and error analysis of a differential rectification method were performed by using digital images taken by a frame camera onboard BILSAT and pushbroom scanner on ASTER. Three methods were implemented for Sensitivity and Uncertainty analysis: Monte Carlo, covariance analysis and FAST (Fourier Amplitude Sensitivity Test). A parameter estimation procedure was carried out on the basis of so called Mixed Model extended by some suitable additional regularization parameters to stabilize the solution for improper geometrical conditions of the imaging system. The effectiveness and accuracy of the differential rectification method were compared with other rectification methods and the results were analyzed. Furthermore the differential method is adapted to the pushbroom scanners and software which provides rectified images from raw satellite images was developed.

No description available.