A novel method for visual place recognition is introduced and evaluated, demonstrating robustness to perceptual aliasing and observation noise. This is achieved by increasing discrimination through a more structured representation of visual observations. Estimation of observation likelihoods are based on graph kernel formulations, utilizing both the structural and visual information encoded in covisibility graphs. The proposed probabilistic model is able to circumvent the typically difficult and expensive posterior normalization procedure by exploiting the information available in visual observations. Furthermore, the place recognition complexity is independent of the size of the map. Results show improvements over the state-of-theart on a diverse set of both public datasets and novel experiments, highlighting the benefit of the approach.

With the growing demand for deployment of robots in real scenarios, robustness in the perception capabilities for navigation lies at the forefront of research interest, as this forms the backbone of robotic autonomy. Existing place recognition approaches traditionally follow the feature-based bag-of-words paradigm in order to cut down on the richness of information in images. As structural information is typically ignored, such methods suffer from perceptual aliasing and reduced recall, due to the ambiguity of observations. In a bid to boost the robustness of appearance-based place recognition, we consider the world as a continuous constellation of visual words, while keeping track of their covisibility in a graph structure. Locations are queried based on their appearance, and modelled by their corresponding cluster of landmarks from the global covisibility graph, which retains important relational information about landmarks. Complexity is reduced by comparing locations by their graphs of visual words in a simplified manner. Test results show increased recall performance and robustness to noisy observations, compared to state-of-the-art methods.

In order to diminish the influence of pose choice during appearance-based mapping, a more natural representation of location models is established using covisibility graphs. As the robot moves through the environment, visual landmarks are detected, and connected if seen as covisible. The introduction of a novel generative model allows relevant subgraphs of the covisibility map to be compared to a given query without needing to normalize over all previously seen locations. The use of probabilistic methods provides a unified framework to incorporate sensor error, perceptual aliasing, decision thresholds, and multiple location matches. The system is evaluated and compared with other state-of-the-art methods.

Camera pose estimation (perspective-n-points, or PnP) is a well-studied problem in computer vision with many applications in robotics. However state-of-art approaches often consider the task of pose estimation with respect to an object where the ratio between the furthest and closest point is small. Localisation in the context of simultaneous localisation and mapping (SLAM) violates this constraint and the naive application of PnP algorithms can lead to biased and imprecise estimates. In this article, we explore weighted object-space errors to provide an efficient, accurate and robust pose estimation solution. State-of-the-art results are shown in realistic scenarios and a practical vision framework is proposed that can be used for visual SLAM. An implementation of the proposed algorithm has been made available as open-source software in Matlab and C++.

Understanding and analysing video data from static or mobile surveillance cameras often requires knowledge of the scene and the camera placement. In this article, we provide a way to simplify the user's task of understanding the scene by rendering the camera view as if observed from the user's perspective by estimating his position using a real-time visual SLAM system. Augmenting the view is referred to as hidden view synthesis. Compared to previous work, the current approach improves by simplifying the setup and requiring minimal user input. This is achieved by building a map of the environment using a visual SLAM system and then registering the surveillance camera in this map. By exploiting the map, a different moving camera can render hidden views in real-time at 30Hz. We discuss some of the challenges remaining for full automation. Results are shown in an indoor environment for surveillance applications and outdoors with application to improved safety in transport.

Though modern Visual Simultaneous Localisation and Mapping (vSLAM) systems are capable of localising robustly and efficiently even in the case of a monocular camera, the maps produced are typically sparse point-clouds that are difficult to interpret and of little use for higher-level reasoning tasks such as scene understanding or human- machine interaction. In this paper we begin to address this deficiency, presenting progress on expanding the competency of visual SLAM systems to build richer maps. Specifically, we concentrate on modelling indoor scenes using semantically meaningful surfaces and accompanying labels, such as “floor”, “wall”, and “ceiling” - an important step towards a representation that can support higher-level reasoning and planning. We leverage the Manhattan world assumption and show how to extract vanishing directions jointly across a video stream. We then propose a guided line detector that utilises known vanishing points to extract extremely subtle axis- aligned edges. We utilise recent advances in single view structure recovery to building geometric scene models and demonstrate our system operating on-line.

We are concerned with enabling truly large scale autonomous navigation in typical human environments. To this end we describe the acquisition and modeling of large urban spaces from data that reflects human sensory input. Over 181GB of image and inertial data are captured using head-mounted stereo cameras. This data is processed into a relative map covering 121 km of Southern England. We point out the numerous challenges we encounter, and highlight in particular the problem of undetected ego-motion, which occurs when the robot finds itself on-or-within a moving frame of reference. In contrast to global-frame representations, we find that the continuous relative representation naturally accommodates moving-reference-frames - without having to identify them first, and without inconsistency. Within a moving-reference-frame, and without drift-less global exteroceptive sensing, motion with respect to the global-frame is effectively unobservable. This underlying truth drives us towards relative topometric solutions like relative bundle adjustment (RBA), which has no problem representing distance and metric Euclidean structure, yet does not suffer inconsistency introduced by the attempt to solve in the global-frame.

This paper proposes a new topo-metric representation of the world based on co-visibility that simplifies data association and improves the performance of appearance-based recognition. We introduce the concept of dynamic bagof-words, which is a novel form of query expansion based on finding cliques in the landmark co-visibility graph. The proposed approach avoids the - often arbitrary - discretisation of space from the robot's trajectory that is common to most image-based loop closure algorithms. Instead we show that reasoning on sets of co-visible landmarks leads to a simple model that out-performs pose-based or view-based approaches. Using real and simulated imagery, we demonstrate that dynamic bag-of-words query expansion can improve precision and recall for appearance-based localisation.

Continuous, real-time mapping of an environment using a camera requires a constant-time estimation engine. This rules out optimal global solving such as bundle adjustment. In this article, we investigate the precision that can be achieved with only local estimation of motion and structure provided by a stereo pair. We introduce a simple but novel representation of the environment in terms of a sequence of relative locations. We demonstrate precise local mapping and easy navigation using the relative map, and importantly show that this can be done without requiring a global minimisation after loop closure. We discuss some of the issues that arise from using a relative representation, and evaluate our system on long sequences processed at a constant 30-45 Hz, obtaining precisions down to a few metres over distances of a few kilometres.

This paper addresses the problem of tracking a planar region of the scene using an uncalibrated omnidirectional camera. Omnidirectional cameras are a popular choice of visual sensors in robotics because the large field of view is well adapted to motion estimation and obstacle avoidance. The novelty of this work resides in simplifying the calibration phase by providing a direct approach to tracking without any prior knowledge of the camera, lens or mirror parameters. We deal with a nonlinear optimization problem that can be solved for small displacements between two images like those acquired at video rate by a camera mounted on a robot. In order to assess the performance of the proposed method, we perform experiments with synthetic and real data.

In this paper we show how online images can be automatically exploited for scene visualization and reconstruction starting from a mere visual query provided by the user. A visual query is used to retrieve images of a landmark place using a visual search engine. These images are used to reconstruct robust 3–D features and camera poses in projective space. Novel views are then rendered corresponding to a virtual camera flying smoothly through the projective space by triangulation of the projected points in the output view. We introduce a method to fuse the rendered novel views from all input images at each virtual view point by computing their intrinsic image and illuminations. This approach allows us to remove the occlusions and maintain consistent and controlled illumination throughout the rendered sequence. We demonstrate the performance of our prototype system on two landmark structures. 1

This article addresses the problem of real-time visual tracking in presence of complex motion blur. Previous authors have observed that efficient tracking can be obtained by matching blurred images instead of applying the computationally expensive task of deblurring (H. Jin et al., 2005). The study was however limited to translational blur. In this work, we analyse the problem of tracking in presence of spatially variant motion blur generated by a planar template. We detail how to model the blur formation and parallelise the blur generation, enabling a real-time GPU implementation. Through the estimation of the camera exposure time, we discuss how tracking initialisation can be improved. Our algorithm is tested on challenging real data with complex motion blur where simple models fail. The benefit of blur estimation is shown for structure and motion.

This paper addresses the problem of motion estimation and 3-D reconstruction through visual tracking with a single-viewpoint sensor and, in particular, how to generalize tracking to calibrated omnidirectional cameras. We analyze different minimization approaches for the intensity-based cost function (sum of squared differences). In particular, we propose novel variants of the efficient second-order minimization (ESM) with better computational complexities and compare these algorithms with the inverse composition (IC) and the hyperplane approximation (HA). Issues regarding the use of the IC and HA for 3-D tracking are discussed. We show that even though an iteration of ESM is computationally more expensive than an iteration of IC, the faster convergence rate makes it globally faster. The tracking algorithm was validated by using an omnidirectional sensor mounted on a mobile robot.

This paper presents a flexible approach for calibrating omnidirectional single viewpoint sensors from planar grids. These sensors are increasingly used in robotics where accurate calibration is often a prerequisite. Current approaches in the field are either based on theoretical properties and do not take into account important factors such as misalignment or camera-lens distortion or over-parametrised which leads to minimisation problems that are difficult to solve. Recent techniques based on polynomial approximations lead to impractical calibration methods. Our model is based on an exact theoretical projection function to which we add well identified parameters to model real-world errors. This leads to a full methodology from the initialisation of the intrinsic parameters to the general calibration. We also discuss the validity of the approach for fish-eye and spherical models. An implementation of the method is available as OpenSource software on the author's Web page. We validate the approach with the calibration of parabolic, hyperbolic, folded mirror, wide-angle and spherical sensors.

International audience

This paper presents several methods for estimating the relative position of a central catadioptric camera (including perspective cameras) and a laser range finder in order to obtain depth information in the panoramic image. The problem is analysed from a robotic perspective and according to the available information (visible or invisible laser beam, partial calibration, drift of laser data,...). The feasibility of the calibration is also discussed. The feature extraction process and real data are presented

This paper presents a parametric approach for tracking piecewise planar scenes with central catadioptric cameras (including perspective cameras). We extend the standard notion of homography to this wider range of devices through the unified projection model on the sphere. We avoid unwarping the image to a perspective view and take into account the non-uniform pixel resolution specific to non-perspective central catadioptric sensors. The homography is parametrised by the Lie algebra of the special linear group SL(3) to ensure that only eight free parameters are estimated. With this model, we use an efficient second-order minimisation technique leading to a fast tracking algorithm with a complexity similar to a first-order approach. The developed algorithm was tested on the estimation of the displacement of a mobile robot in a real application and proved to be very precise

In this paper we present an analysis of 3D line projections for central catadioptric cameras from a projective perspective. Most algorithms consider the projection of lines as general conics in the image plane with five degrees of freedom. However, in the calibrated case, only two parameters are needed to represent lines. We describe methods to obtain fast extraction and estimation algorithms. We then explain how classical edge-tracking algorithms can be adapted to these sensors. To this avail, we introduce two parametric equations for lines in central catadioptric images. We then propose a minimal representation for the euclidean transformation in the structure from motion problem and introduce possible metrics between a point and a central catadioptric line. These metrics are evaluated on simulated data. The structure from motion algorithm, from the line extraction process to the tracking and the reconstruction is tested on a real sequence