Previous publications have demonstrated, inter alia, the generation of saccades where the gaze direction of the system is rapidly redirected to a target of interest, followed by periods of smooth pursuit where fixation is maintained upon a target [6], [3]. The gaze controller, implemented as a finite state machine, selects and reselects these behaviours to provide robust performance in natural dynamic scenes over extended periods. One of the housekeeping operations of the servo-controller is to propagate odometry — in particular the axis positions or gaze angles  ¢ ¡ and  ¤ £ — around the system. For an extended pursuit sequence, the sequence ¥ ¥¨§�©� � ¥�§�©� © of gaze angles for the platform’s elevation axis and one vergence axis defines an observer-based trajectory for the target. Our first aim here is to recover an observer trajectory as the camera pursues an object. Various algorithms have been implemented to drive pursuit behaviour using active cameras, such as correlation [20], [21], deformable templates [22], feature-based affine transfer [4], [5] and segmented optical flow [6]. The last is used here, though any of the variety of methods might be employed to equal effect. Optical flow is recovered in a small central, or foveal, area of the image. Because of its small

Traffic statistics desired by road engineers and planners, and "traffic warning" systems demand real-time performance which precludes the use of batch processing. We apply recent real-time tracking techniques along with scene specific tuning of the dynamics to enable the tracker to accurately predict target location and thus reduce the amount of search and/or image processing required. The benefits of learning dynamics for accurate prediction are speed -- our tracker operates at frame rate -- and smoothing of vibration. Initial calibration of the projective relationship between the image and ground planes enables metric information to be derived from the image positions and velocities without full camera calibration. Results are presented on real-world traffic scenes showing the tracker to be both fast and robust to vibrations which are inevitable in traffic locations. 1 Introduction and Related Work Automatic surveillance of traffic has been of interest for many years. As early as ...

Localization and navigation with sensory feedback are two main issues in the area of mobile robot research [1, 2, 3, 4]. Many papers have addressed this problem, in which all kinds of sensors are used (e.g. cameras, sonars, lightstripers, touch sensors, etc.). In this context, the particular configuration of a robot is often represented with cartesian coordinates (x; y) and a heading angle OE. Though this might be a very convenient way to handle positioning, especially in a global map, an alternative representation may be useful for understanding the relationship between a sensor and its environment. In this paper, we propose a representation that uses angles to describe the position of a camera with respect to an object (more precisely, two points of an object). It is based on the visual angle and the range ratio. The representation is locally cartesian and better reflects the visual perception. The second part of the paper is dedicated to the analyze of the groundplane constraint and...

Introduction When moving through the environment one experiences a specific visual impression. The visual field seems to flow towards and around the observer. The flow starts at the point towards which one is heading. The first to describe this phenomenon was Helmholtz [1] in the second half of last century. Later, Gibson [2, 3, 4] introduced the concept of the optical flow field, an array of vectors which represent the velocity of picture elements between successive images. Since Gibsons non--mathematical work many technical studies have been devoted to mathematical modeling of the process of 3D motion perception. In principle it is possible to infer useful information about the spatial layout of the scene, relative motion parameters and hence for the controlling of one's own movement [5]. Possible collisions can be recognized early on since approaching objects result in relatively high retinal velocities. This is vital for animals moving fast in narrow habitat