In this thesis, we develop an algorithm for detecting independent motion in real-time from 2D image sequences of arbitrarily complex 3D scenes. We discuss the necessary background information in image formation, optical flow, multiple view geometry, robust estimation, and real-time camera and scene pose estimation for constructing and understanding the operation of our algorithm. Furthermore, we provide an overview of existing independent motion detection techniques and compare them to our proposed solution. Unfortunately, the existing independent motion detection techniques were not evaluated quantitatively nor were their source code made publicly available. Therefore, it is not possible to make direct comparisons. Instead, we constructed several comparison algorithms which should have comparable performance to these previous approaches. We developed methods for quantitatively comparing independent motion detection algorithms and found that our solution had the best performance. By establishing a method for quantitatively evaluating these algorithms and publishing our results, we hope to foster better research in this area and help future investigators more quickly advance the state of the art. / text

This dissertation deals with the image-based approach to synthesize a virtual scene using sparse images or a video sequence without the use of 3D models. In our scenario, a real dynamic or static scene is captured by a set of un-calibrated images from different viewpoints. After automatically recovering the geometric transformations between these images, a series of photo-realistic virtual views can be rendered and a virtual environment covered by these several static cameras can be synthesized. This image-based approach has applications in object recognition, object transfer, video synthesis and video compression. In this dissertation, I have contributed to several sub-problems related to image based view synthesis. Before image-based view synthesis can be performed, images need to be segmented into individual objects. Assuming that a scene can approximately be described by multiple planar regions, I have developed a robust and novel approach to automatically extract a set of affine or projective transformations induced by these regions, correctly detect the occlusion pixels over multiple consecutive frames, and accurately segment the scene into several motion layers. First, a number of seed regions using correspondences in two frames are determined, and the seed regions are expanded and outliers are rejected employing the graph cuts method integrated with level set representation. Next, these initial regions are merged into several initial layers according to the motion similarity. Third, the occlusion order constraints on multiple frames are explored, which guarantee that the occlusion area increases with the temporal order in a short period and effectively maintains segmentation consistency over multiple consecutive frames. Then the correct layer segmentation is obtained by using a graph cuts algorithm, and the occlusions between the overlapping layers are explicitly determined. Several experimental results are demonstrated to show that our approach is effective and robust. Recovering the geometrical transformations among images of a scene is a prerequisite step for image-based view synthesis. I have developed a wide baseline matching algorithm to identify the correspondences between two un-calibrated images, and to further determine the geometric relationship between images, such as epipolar geometry or projective transformation. In our approach, a set of salient features, edge-corners, are detected to provide robust and consistent matching primitives. Then, based on the Singular Value Decomposition (SVD) of an affine matrix, we effectively quantize the search space into two independent subspaces for rotation angle and scaling factor, and then we use a two-stage affine matching algorithm to obtain robust matches between these two frames. The experimental results on a number of wide baseline images strongly demonstrate that our matching method outperforms the state-of-art algorithms even under the significant camera motion, illumination variation, occlusion, and self-similarity. Given the wide baseline matches among images I have developed a novel method for Dynamic view morphing. Dynamic view morphing deals with the scenes containing moving objects in presence of camera motion. The objects can be rigid or non-rigid, each of them can move in any orientation or direction. The proposed method can generate a series of continuous and physically accurate intermediate views from only two reference images without any knowledge about 3D. The procedure consists of three steps: segmentation, morphing and post-warping. Given a boundary connection constraint, the source and target scenes are segmented into several layers for morphing. Based on the decomposition of affine transformation between corresponding points, we uniquely determine a physically correct path for post-warping by the least distortion method. I have successfully generalized the dynamic scene synthesis problem from the simple scene with only rotation to the dynamic scene containing non-rigid objects. My method can handle dynamic rigid or non-rigid objects, including complicated objects such as humans. Finally, I have also developed a novel algorithm for tri-view morphing. This is an efficient image-based method to navigate a scene based on only three wide-baseline un-calibrated images without the explicit use of a 3D model. After automatically recovering corresponding points between each pair of images using our wide baseline matching method, an accurate trifocal plane is extracted from the trifocal tensor implied in these three images. Next, employing a trinocular-stereo algorithm and barycentric blending technique, we generate an arbitrary novel view to navigate the scene in a 2D space. Furthermore, after self-calibration of the cameras, a 3D model can also be correctly augmented into this virtual environment synthesized by the tri-view morphing algorithm. We have applied our view morphing framework to several interesting applications: 4D video synthesis, automatic target recognition, multi-view morphing.

Multiple view geometry is the foundation of an important class of computer vision techniques for simultaneous recovery of camera motion and scene structure from a set of images. There are numerous important applications in this area. Examples include video post-production, scene reconstruction, registration, surveillance, tracking, and segmentation. In video post-production, which is the topic being addressed in this dissertation, computer analysis of the motion of the camera can replace the currently used manual methods for correctly aligning an artificially inserted object in a scene. However, existing single view methods typically require multiple vanishing points, and therefore would fail when only one vanishing point is available. In addition, current multiple view techniques, making use of either epipolar geometry or trifocal tensor, do not exploit fully the properties of constant or known camera motion. Finally, there does not exist a general solution to the problem of synchronization of N video sequences of distinct general scenes captured by cameras undergoing similar ego-motions, which is the necessary step for video post-production among different input videos. This dissertation proposes several advancements that overcome these limitations. These advancements are used to develop an efficient framework for video analysis and post-production in multiple cameras. In the first part of the dissertation, the novel inter-image constraints are introduced that are particularly useful for scenes where minimal information is available. This result extends the current state-of-the-art in single view geometry techniques to situations where only one vanishing point is available. The property of constant or known camera motion is also described in this dissertation for applications such as calibration of a network of cameras in video surveillance systems, and Euclidean reconstruction from turn-table image sequences in the presence of zoom and focus. We then propose a new framework for the estimation and alignment of camera motions, including both simple (panning, tracking and zooming) and complex (e.g. hand-held) camera motions. Accuracy of these results is demonstrated by applying our approach to video post-production applications such as video cut-and-paste and shadow synthesis. As realistic image-based rendering problems, these applications require extreme accuracy in the estimation of camera geometry, the position and the orientation of the light source, and the photometric properties of the resulting cast shadows. In each case, the theoretical results are fully supported and illustrated by both numerical simulations and thorough experimentation on real data.

This thesis is concerned with the utilization of object symmety as a cue for segmentation and object recognition. In particular it investigates the problem of detecting 3D bilaterally symmetric objects from affine views. The first part of the thesis investigates the problem of detecting 3D bilateral symmetry within a scene from known point correspondences across two or more affine views. We begin by extending the notion of skewed symmetry to three dimensions, and give a definition in terms of degenerate structure that applies equally to an affine 3D structure or to point correspondences across two or more affine views. We then consider the effects of measurement errors on symmetry detection, and derive an optimal statistical test of degenerate structure, and thereby of 3D-skewed symmetry. We then move on to the problem of searching for 3D skewed symmetric sets within a larger scene. We discuss two approaches to the problem, both of which we have implemented, and we demonstrate fully automatic detection of 3D skewed symmetry on images of uncluttered scenes. We conclude the first part by investing means of verifying the presence of bilateral rather than skewed symmetry in the Euclidean space, by enforcing mutual consistency between multiple skewed symmetric sets, and by drawing on partial knowledge about the camera calibration. The second part of the thesis is concerned with the problem of obtaining feature correspondences across multiple affine views, as required for the detection of symmetry. In particular we investigate the geometric matching constraints that exist between affine views. We start by specilizing the four projective multifocal tensors to the affine case, and use these to carry the bulk of all known projective multi-view matching relations to affine views, unearthing some new relations in the process. Having done that, we address the problem of estimating the affine tensors. We provide a minimal set of constraints on the affine trifocal tensor, and search for ways of estimating the affine tensors from point and line correspondences.

This is a thesis on outdoor monocular visual SLAM in natural environments. The techniques proposed herein aim at estimating camera pose and 3D geometrical structure of the surrounding environment. This problem statement was motivated by the GPS-denied scenario for a sea-surface vehicle developed at Plymouth University named Springer. The algorithms proposed in this thesis are mainly adapted for the Springer’s environmental conditions, so that the vehicle can navigate on a vision based localization system when GPS is not available; such environments include estuarine areas, forests and the occasional semi-urban territories. The research objectives are constrained versions of the ever-abiding problems in the fields of multiple view geometry and mobile robotics. The research is proposing new techniques or improving existing ones for problems such as scene reconstruction, relative camera pose recovery and filtering, always in the context of the aforementioned landscapes (i.e., rivers, forests, etc.). Although visual tracking is paramount for the generation of data point correspondences, this thesis focuses primarily on the geometric aspect of the problem as well as with the probabilistic framework in which the optimization of pose and structure estimates takes place. Besides algorithms, the deliverables of this research should include the respective implementations and test data for these algorithms in the form of a software library and a dataset containing footage of estuarine regions taken from a boat, along with synchronized sensor logs. This thesis is not the final analysis on vision based navigation. It merely proposes various solutions for the localization problem of a vehicle navigating in natural environments either on land or on the surface of the water. Although these solutions can be used to provide position and orientation estimates when GPS is not available, they have limitations and there is still a vast new world of ideas to be explored.