Additionally, we suggest a new formulation for the perspective camera projection matrix. In particular, regarding how the 3 x 3 rotation matrix, R, of the camera should be incorporated into the 3 x 4 camera projection matrix, P. We show that the incorporated rotation should neither be the camera rotation R nor its transpose, but a reversed (left-handed) version of it. The fundamental matrix between a pair of stereo cameras is reformulated more accurately accordingly. This is extremely useful when we want to calculate the fundamental matrix accurately from the stereo camera matrices. It is especially true when the feature correspondences are too few for robust methods, such as RANSAC, to operate. We expect that this new model would have an impact on various applications. / Furthermore, the process of estimating the rotation and translation parameters between a stereo pair from the essential matrix is investigated. This is an essential step for our multi-camera pose estimation method. We show that there are 16 solutions based on the singular value decomposition (not four or eight as previously thought). We also suggest a checking step to ascertain that the proposed algorithm will come up with accurate results. The checking step ensures the accuracy of the fundamental matrix calculated using the pose obtained. This provides a speedy way to calibrate a stereo rig. Our proposed theories are supported by the real and synthetic data experiments reported in this thesis. / In this thesis, we solve the pose estimation problem for robot motion by placing multiple cameras on the robot. In particular, we use four cameras arranged as two back-to-back stereo pairs combined with the Extended Kalman Filter (EKF). The EKF is used to provide a frame by frame recursive solution suitable for the real-time application at hand. The reason for using multiple cameras is that the pose estimation problem is more constrained for multiple cameras than for a single camera. Their use is further justified by the drop in price which is accompanied by the remarkable increase in accuracy. Back-to-back cameras are used since they are likely to have a larger field of view, provide more information, and capture more features. In this way, they are more likely to disambiguate the pose translation and rotation parameters. Stereo information is used in self-initialization and outlier rejection. Simple yet efficient methods have been proposed to tackle the problem of long-sequence drift. Our approaches have been compared, under different motion patterns, to other methods in the literature which use a single camera. Both the simulations and the real experiments show that our approaches are the most robust and accurate among them all as well as fast enough to realize the real-time requirement of robot navigation. / Mohammad Ehab Mohammad Ragab. / "April 2008." / Adviser: K. H. Wong. / Source: Dissertation Abstracts International, Volume: 70-03, Section: B, page: 1763. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2008. / Includes bibliographical references (p. 138-148) and index. / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. [Ann Arbor, MI] : ProQuest Information and Learning, [200-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstracts in English and Chinese. / School code: 1307.

Thesis (M.Sc.)--York University, 2006. Graduate Programme in Computer Science. / Typescript. Includes bibliographical references (leaves 157-162). Also available on the Internet. MODE OF ACCESS via web browser by entering the following URL: http://gateway.proquest.com/openurl?url_ver=Z39.88-2004&res_dat=xri:pqdiss&rft_val_fmt=info:ofi/fmt:kev:mtx:dissertation&rft_dat=xri:pqdiss:MR29596

In this work we focus on building local maps to enable wheeled mobile robots to navigate safely and autonomously in urban environments. Urban environments present a variety of hazards that mobile robots have to detect and represent in their maps to navigate safely. Examples of hazards include obstacles such as furniture, drop-offs such as at downward stairs, and inclined surfaces such as wheelchair ramps. We address two shortcomings perceived in the literature on mapping. The first is the extensive use of expensive laser-based sensors for mapping, and the second is the focus on only detecting obstacles when clearly other hazards such as drop-offs need to be detected to ensure safety. Therefore, in this work we develop algorithms for building maps using only relatively inexpensive stereo cameras, that allow safe local navigation by detecting and modeling hazards such as overhangs, drop-offs, and ramps in addition to static obstacles. The hazards are represented using 2D annotated grid maps called local safety maps. Each cell in the map is annotated with one of several labels: Level, Inclined, Non-ground, or, Unknown. Level cells are safe for travel whereas Inclined cells require caution. Non-ground cells are unsafe for travel and represent obstacles, overhangs, or regions lower than safe ground. Level and Inclined cells can be further annotated as being Drop-off Edges. The process of building safety maps consists of three main steps: (i) computing a stereo depth map; (ii) building a 3D model using the stereo depths; and, (iii) analyzing the 3D model for safety to construct the safety map. We make significant contributions to each of the three steps: we develop global stereo methods for computing disparity maps that use edge and color information; we introduce a probabilistic data association method for building 3D models using stereo range points; and we devise a novel method for segmenting and fitting planes to 3D models allowing for a precise safety analysis. In addition, we also develop a stand-alone method for detecting drop-offs in front of the robot that uses motion and occlusion cues and only relies on monocular images. We introduce an evaluation framework for evaluating (and comparing) our algorithms on real world data sets, collected by driving a robot in various environments. Accuracy is measured by comparing the constructed safety maps against ground truth safety maps and computing error rates. The ground truth maps are obtained by manually annotating maps built using laser data. As part of the framework we also estimate latencies introduced by our algorithms and the accuracy of the plane fitting process. We believe this framework can be used for comparing the performance of a variety of vision-based mapping systems and for this purpose we make our datasets, ground truth maps, and evaluation code publicly available. We also implement a real-time version of one of the safety map algorithms on a wheelchair robot and demonstrate it working in various environments. The constructed safety maps allow safe local motion planning and also support the extraction of local topological structures that can be used to build global maps. / text

published_or_final_version / Electrical and Electronic Engineering / Master / Master of Philosophy

To improve the accuracy in capturing the motion of deformable objects, a reconfigurable multi-camera active-vision system which can dynamically reposition its cameras online is proposed, and a design for such a system, along with a methodology to select the near-optimal positions and orientations of the set of cameras, is presented. The active-vision system accounts for the deformation of the object-of-interest by tracking triangulated vertices in order to predict the shape of the object at subsequent demand instants. It then selects a system configuration that minimizes the expected error in the recovered position of each of these vertices. Extensive simulations and experiments have verified that using the proposed reconfigurable system to both translate and rotate cameras to near-optimal poses is tangibly superior to using cameras which are either static, or can only rotate, in minimizing the error in recovered vertex positions.

To improve the accuracy in capturing the motion of deformable objects, a reconfigurable multi-camera active-vision system which can dynamically reposition its cameras online is proposed, and a design for such a system, along with a methodology to select the near-optimal positions and orientations of the set of cameras, is presented. The active-vision system accounts for the deformation of the object-of-interest by tracking triangulated vertices in order to predict the shape of the object at subsequent demand instants. It then selects a system configuration that minimizes the expected error in the recovered position of each of these vertices. Extensive simulations and experiments have verified that using the proposed reconfigurable system to both translate and rotate cameras to near-optimal poses is tangibly superior to using cameras which are either static, or can only rotate, in minimizing the error in recovered vertex positions.

No description available.

No description available.

No description available.

No description available.