Monocular SLAM for Visual Odometry
ABSTRACT The ego-motion online estimation process from a video input is often called visual odometry. Typically optical flow and structure from motion (SFM) techniques have been used for visual odometry. Monocular simultaneous localization and mapping (SLAM) techniques implicitly estimate camera ego-motion while incrementally build a map of the environment. However in monocular SLAM, when the number of features in the system state increases, the computational cost grows rapidly; consequently maintaining frame rate operation becomes impractical. In this paper monocular SLAM is proposed for map-based visual odometry. The number of features is bounded removing features dynamically from the system state, for maintaining a stable processing time. In the other hand if features are removed then previous visited sites can not be recognized, nevertheless in an odometry context this could not be a problem. A method for feature initialization and a simple method for recovery metric scale are proposed. The experimental results using real image sequences show that the scheme presented in this paper is promising.
- SourceAvailable from: Antoni Grau
[Show abstract] [Hide abstract]
- "La búsqueda de las características se limita a regiones suscritas alrededor de la predicción h i y con área definida por la matriz de innovación (30) donde H i es el Jacobiano del modelo de observación con respecto al estado x̂ , P k es la matriz de covariancias a priori del estado. Se suponen mediciones corrompidas por ruido Gaussiano con media cero y covarianza R. Si desea más detalles acerca de la adaptación del método MTE para su aplicación en SLAM monocular, consulte Munguía y Grau (2007). Con el fin de validar experimentalmente el método MTE (en su variante de aplicación al SLAM "
ABSTRACT: The SLAM or Simultaneous Localization and Mapping, is a technique in which a robot or autonomous vehicle operate in an a priori unknown environment, using only its onboard sensors to simultaneously build a map of its surroundings and use it to track its position. The sensors have a large impact on the algorithm used for SLAM. Recent approaches are focusing on the use of cameras as the main sensor, because they yield a lot of information and are well adapted for embedded systems: they are light, cheap and power saving. However, unlike range sensors which provide range and angular information, a camera is a projective sensor which measures the bearing of images features. Therefore depth information (range) cannot be obtained in a single step. This fact has propitiated the emergence of a new family of SLAM algorithms: the Bearing-Only SLAM methods, which mainly rely in especial techniques for features system-initialization in order to enable the use of bearing sensors (as cameras) in SLAM systems. In this work a practical method is presented, for initializing new features in bearing-only SLAM systems. The proposed method, defines a single hypothesis for the initial depth of features, by the use of an stochastic technique of triangulation. Several simulations as well two scenarios of applications with real data are presented, in order to show the performance of the proposed method.06/2013; 14(2):257-274. DOI:10.1016/S1405-7743(13)72241-8
- "Various approaches to resolve the scale ambiguity and to establish a metric coordinate system have been proposed. For example, Davison et al.  add a calibration object with known size to the scene, or Munguia et al.  require manual selection of three scene points with known distance. Likewise, if objects with known size can be identified, they can be used to determine the metric scale. "
Conference Paper: Scale-preserving long-term visual odometry for indoor navigation[Show abstract] [Hide abstract]
ABSTRACT: We present a visual odometry system for indoor navigation with a focus on long-term robustness and consistency. As our work is targeting mobile phones, we employ monocular SLAM to jointly estimate a local map and the device's trajectory. We specifically address the problem of estimating the scale factor of both, the map and the trajectory. State-of-the-art solutions approach this problem with an Extended Kalman Filter (EKF), which estimates the scale by fusing inertial and visual data, but strongly relies on good initialization and takes time to converge. Each visual tracking failure introduces a new arbitrary scale factor, forcing the filter to re-converge. We propose a fast and robust method for scale initialization that exploits basic geometric properties of the learned local map. Using random projections, we efficiently compute geometric properties from the feature point cloud produced by the visual SLAM system. From these properties (e.g., corridor width or height) we estimate scale changes caused by tracking failures and update the EKF accordingly. As a result, previously achieved convergence is preserved despite re-initializations of the map. To minimize the time required to continue tracking after failure, we perform recovery and re-initialization in parallel. This increases the time available for recovery and hence the likelihood for success, thus allowing almost seamless tracking. Moreover, fewer re-initializations are necessary. We evaluate our approach using extensive and diverse indoor datasets. Results demonstrate that errors and convergence times for scale estimation are considerably reduced, thus ensuring consistent and accurate scale estimation. This enables long-term odometry despite of tracking failures which are inevitable in realistic scenarios.2012 International Conference on Indoor Positioning and Indoor Navigation (IPIN); 01/2012
[Show abstract] [Hide abstract]
- "To maintain the computational cost bounded (a requirement for real-time VIO), features that leave the field of view of the camera can be removed from the state vector . On the other hand, EKF algorithms exist that only maintain a sliding window of camera poses in the state vector, and use the feature observations to apply probabilistic constraints between these poses (e.g., , ). "
ABSTRACT: In this paper, we perform a rigorous analysis of EKF-based visual-inertial odometry (VIO) and present a method for improving its performance. Specifically, we examine the properties of EKF-based VIO, and show that the standard way of computing Jacobians in the filter inevitably causes inconsistency and loss of accuracy. This result is derived based on an observability analysis of the EKF's linearized system model, which proves that the yaw erroneously appears to be observable. In order to address this problem, we propose modifications to the multi-state constraint Kalman filter (MSCKF) algorithm , which ensure the correct observability properties without incurring additional computational cost. Extensive simulation tests and real-world experiments demonstrate that the modified MSCKF algorithm outperforms competing methods, both in terms of consistency and accuracy.Proceedings - IEEE International Conference on Robotics and Automation 01/2012; DOI:10.1109/ICRA.2012.6225229