Real-Time Quantized Optical Flow

Camus, T.

doi:10.1006/rtim.1996.0048

Cited by 51 publications

(5 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Given that the current image size allows for about a 20% idle time to buffer operating system (OS) events, it is likely that a system making use of realtime OS facilities could use this available CPU time to process a larger image. It has been argued that there are computational advantages in keeping the search radius of the optical flow algorithm as small as one pixel [6] and keep the frame rate high.…”

Section: Discussionmentioning

confidence: 99%

“…Although this is sufficient for various robotics vision tasks [6][10] [ 113, it is not sufficient for our application since the calculation of divergence requires that the spatial derivatives of the optical flow can be measured. Because the quantized optical flow is basically a step function, these derivatives do not exist.…”

Section: Full Image Flow Estimationmentioning

confidence: 99%

“…Figure 3 shows this graphically; thin lines represent parabolic interpolations of three given correlation match values. This double one-dimensional interpolation calculation cuts the frame rate approximately in half: real-valued optical flow can be computed on 32x64 images, calculating and such templates may be performed extremely quickly using a boxjilter as described in [6]. In order to improve the consistency of the divergence estimates, we apply temporal and spatial median filters to the individual divergence calculations.…”

Section: Y1 -Y O -A ( ( X 1 ) 2 -( X O ) 2 ) B = X I -Xomentioning

confidence: 99%

See 2 more Smart Citations

Real-time single-workstation obstacle avoidance using only wide-field flow divergence

Camus

Coombs

Herman

et al. 1996

Proceedings of 13th International Conference on Pattern Recognition

View full text Add to dashboard Cite

A real-time robot vision system is described which uses only the divergence of the optical $ow field for both steering control and collision detection. The robot has wandered about the lab at 20 c d s for as long as 26 minutes without collision. The entire system is implemented on a single ordinary UNIX workstation without the benefit of real-time operating system support. Dense opticaljlow data are calculated in real-time across the entire wide-angle image. The divergence of this optical $ow field is calculated everywhere and used to control steering and collision behaviol: Divergence alone has proven sufJicient for steering past objects and detecting imminent collision. The m j o r contribution is the demonstration of a simple, robust, minimal system that uses $ow-derived measures to control steering and speed to avoid collision in real time for extended periods. Such a system can be embedded in a general, multi-level perceptiodcontrol system.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Full Image Flow Estimationmentioning

confidence: 99%

Section: Y1 -Y O -A ( ( X 1 ) 2 -( X O ) 2 ) B = X I -Xomentioning

confidence: 99%

See 1 more Smart Citation

Real-time single-workstation obstacle avoidance using only wide-field flow divergence

Camus

Coombs

Herman

et al. 1996

Proceedings of 13th International Conference on Pattern Recognition

View full text Add to dashboard Cite

show abstract

“…Theoretical analyses have indicated that optic flow information is sufficient to inform people, animals and machines of the spatial layout of their environment and the direction and heading of themselves and moving or stationary objects [4][5][6]. These formal mathematical analyses have been important in developing artificial vision algorithms for navigation as well as strategies for the compression of video images [7][8][9]. Physiological studies have indicated neural areas that appear to be specialized for retrieving optic flow information in insects [10][11][12], primates [13][14][15][16][17] and humans [18][19][20].…”

Section: Introductionmentioning

confidence: 99%

Effect of prism orientation on heading detection in optic flow

Odom

Ghude

Humble

2006

Journal of Modern Optics

View full text Add to dashboard Cite

Recent studies of ambulatory behaviour indicate that people follow a curvilinear path when moving toward a target and wearing a prism. These results have been interpreted to indicate that heading judgments are based on moving toward a target not on optic flow. This interpretation assumes that if heading judgments were based on optic flow information they would not be altered by prism alignment. We tested this assumption. Thirteen observers made heading judgments using a relatively standard optic flow display. Observers were seated 50 cm from a display that subtended 44.6 arc deg in the horizontal dimension. The observers' task was to indicate if the focus of expansion (FoE) was to the right or left of centre. A double staircase was employed to determine right and left thresholds for the FoEs. Bias was the mean of the right and left thresholds. Precision was 1=2 the difference of the right and left thresholds. Observers repeated the judgments 3 times under 3 conditions: no prism and 2 prism conditions (15 diopters binocular, yoked base right and left). A 3 Â 3 Latin square design was used to avoid order effects. Wearing prisms had a statistically significant effect on subjects' bias ( p < 0:05) without a statistically significant effect on their precision ( p > 0.10). As prisms changed from 15 base left to 15 base right there was a shift to the right in the bias of observers' judgments of about 38 arc min ( p < 0.05). 11 of the 13 observers showed this change. The precision of judgments of heading was unaffected. The difference between the precision of base left and right judgments averaged about 6 arc min. Prism orientation affects the bias of heading judgments in optic flow fields. Precision of heading judgments is unaffected.

show abstract

“…We have been experimenting with some tracking applications, namely XVision (Hager & Toyama 1998) and have been able to obtain frame rates for tracking blob and SSD regions of 20 Hz with only 16.2% CPU usuage and 8% memory usuage on a 233 Mhz Pentium 2 machine using the Big Picture camera. Our plans are to develop coarse depth from motion parallax techniques (Camus 1997) based on selective image regions (i.e., focus of attention) (Reisfeld, Wolfson, & Yeshurun 1995).…”

Section: Map Updates and Localizationmentioning

confidence: 99%

Dynamic issues for mobile robot real-time discovery and path planning

Zelek

Proceedings 1999 IEEE International Symposium on Computational Intelligence in Robotics and Automation. CIRA'99 (Cat. No.99EX37

View full text Add to dashboard Cite

Sensor-based discovery path planning is the on-line guidance of a discovery agent (i.e., robot) who begins execution without a complete a priori map. The presented approach is locally optimal in the sense of minimizing the hitting probability while minimizing the distance to the goal based on what is currently known about the world. The technique is implemented as a process-based client-server. The integration and relationship between different facets of the algorithm are explored in terms of their affect on the real-time performance of the robot. The computation time for the minimal hitting probability surface is reduced by using methods of relaxation and a non-uniform sampling grid. The path is executed by a separate process reading gradient values off of the computed surface at regular intervals. Sensor updates are input as a collection of geometric primitives using a simple interface. Perceptual processing of sensor data is accomplished in separate processes. The algorithm has been implemented and is currently being tested on a'Nomadics Scout platform. Current work is focused on incorporating depth information from passive visual sensors.Keywords-autonomous mobile robot, navigation, dynamic path planning, harmonic functions 0-7803-5806-6/99/$10.00 0 1999 IEEE

show abstract

Real-Time Quantized Optical Flow

Cited by 51 publications

References 29 publications

Real-time single-workstation obstacle avoidance using only wide-field flow divergence

Real-time single-workstation obstacle avoidance using only wide-field flow divergence

Effect of prism orientation on heading detection in optic flow

Dynamic issues for mobile robot real-time discovery and path planning

Contact Info

Product

Resources

About