• Corpus ID: 16207818

Vehicle Tracking at Urban Intersections Using Dense Stereo

  title={Vehicle Tracking at Urban Intersections Using Dense Stereo},
  author={Alexander Barth and David Pfeiffer and Uwe Franke},
A new approach for vehicle tracking at urban intersections based on stereo-vision is proposed. Objects are represented as rigid 3D point clouds and tracked by means of extended Kalman filtering. In this contribution we combine the advantages of a generic feature-based 3D point cloud model with vehicle specific geometrical and kinematical constraints to estimate the pose and motion state of oncoming vehicles at intersections. Real-time dense stereo disparity maps provide new opportunities in… 

Figures from this paper

Tracking oncoming and turning vehicles at intersections
  • A. BarthUwe Franke
  • Computer Science
    13th International IEEE Conference on Intelligent Transportation Systems
  • 2010
A real-time multi-filter approach for vehicle tracking at intersections with a gauge consistency criteria as well as a robust outlier detection method allow for dealing with sudden accelerations and self-occlusions during turn maneuvers.
Moving vehicle detection by optimal segmentation of the Dynamic Stixel World
This paper focuses on the segmentation of space-time data obtained from stereo image sequences, and shows that the optimal segmentation can be efficiently found by means of dynamic programming, for an arbitrary number of objects in the scene.
Robust and Precise 3 D-Modeling of Traffic Scenes based on Dense Stereo Vision
This work proposes so called “stixels” to represent the 3D scene efficiently and shows that the inherent spatial integration delivers depth information with an unprecedented accuracy.
First stereo video dataset with ground truth for remote car pose estimation using satellite markers
A marker-based technique is employed for creating ground truth of car pose and a dataset of a moving vehicle collected from a static mounted stereo camera is created, which serves as a test bed for car pose estimation algorithms.
A Survey of Vision-Based Traffic Monitoring of Road Intersections
This paper provides an overview of vehicle perception systems at road intersections and representative related data sets, and presents possible research directions that are likely to improve the performance of vehicle detection and tracking at intersections.
Ground truth evaluation of the Stixel representation using laser scanners
This paper analyzes the precision of the proposed “Stixel World” representation using a high performance laser scanner as reference sensor and confirms the high accuracy as expected from visual inspection.
Modeling Dynamic 3D Environments by Means of The Stixel World
A medium-level representation named Stixel World, which is a compact but flexible representation of the three-dimensional traffic situation, and the underlying model assumption is that objects stand on the ground and have a vertical pose with a flat surface.
Pixels, Stixels, and Objects
This paper shows how the medium-level representation named Stixel World forms the foundation for a very efficient, robust and comprehensive understanding of traffic scenes.
Integer Occupancy Grids : a probabilistic multi-sensor fusion framework for embedded perception. (Grille d'occupation entière : une méthode probabiliste de fusion multi-capteurs pour la perception embarquée)
Integer Occupancy Grids enable a real-time computation of multi-sensor fusion on embedded low-cost and low-power processing platforms dedicated for automotive applications.
Situationsklassifikation und Bewegungsprognose in Verkehrssituationen mit mehreren Fahrzeugen
Fahrerassistenzsysteme sind in der heutigen Zeit einer der erfolgversprechendsten Beitrage zu mehr Sicherheit auf unseren Strasen. Die intelligenten Helferlein unterstutzen den Fahrer aktiv in


Stereovision Based Vehicle Tracking in Urban Traffic Environments
The corner of the cuboid becomes the central working concept, thus improving the handling of partially occluded objects, of objects partially out of the field of view, and of objects whose measurement is fragmented by the sensor inaccuracies.
3D pose estimation of vehicles using a stereo camera
The results show that the proposed polar distance metric yields a higher accuracy for yaw angle estimation of vehicles than the common Euclidean distance metric, especially when using pixel-accurate stereo points.
Where will the oncoming vehicle be the next second?
A new image based approach for fast and robust tracking of vehicles from a moving platform is presented, which combines the knowledge of where a point in the rigid point cloud has moved within a given time interval with the dynamic model of a vehicle.
Dynamic 3D Scene Analysis from a Moving Vehicle
A system that integrates fully automatic scene geometry estimation, 2D object detection, 3D localization, trajectory estimation, and tracking for dynamic scene interpretation from a moving vehicle and demonstrates the performance of this integrated system on challenging real-world data showing car passages through crowded city areas.
Initialization of Model-Based Vehicle Tracking in Video Sequences of Inner-City Intersections
A fully automatic initialization approach for 3D-model-based vehicle tracking has been developed, based on Edge-Element and Optical-Flow association, which achieves results comparable to those obtained by earlier experiments based on semi-interactive initialization.
6D-Vision: Fusion of Stereo and Motion for Robust Environment Perception
A rich 6D representation is presented that allows the detection of moving obstacles even in the presence of partial occlusion of foreground or background and a powerful fusion of depth and motion information for image sequences taken from a moving observer is presented.
A Robust Approach for Ego-Motion Estimation Using a Mobile Stereo Platform
A robust approach to the problem of ego-motion estimation using a mobile stereo platform, computed for every frame obtaining 3D points of the environment, and the integration of successive movements allows the reconstruction of the travelled path.
B-spline modeling of road surfaces for freespace estimation
A general technique for robust modeling of non-planar ground surfaces based on representing the ground plane as a parametric B-Spline surface and on robust tracking of the surface parameters using stereo disparities and a Kalman filter is proposed.
A real-time computer vision system for measuring traffic parameters
This paper describes the feature-based tracking approach for the task of tracking vehicles under congestion, a real-time implementation using a network of DSP chips, and experiments of the system on approximately 44 lane hours of video data.
Stereo-based Free Space Computation in Complex Traffic Scenarios
This paper presents a method for the computation of free space in complex traffic scenarios by integrating stereo disparity images over time and dynamic programming on a polar-like occupancy grid yields the free space.