Advances in Model-Based Traffic Vision

Worrall, Anthony D.; Sullivan, Geoffrey D.; Baker, Keith D.

doi:10.5244/c.7.56

Cited by 22 publications

(4 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It operates by evaluating evidence for straight lines in one or more images of an object [20]. A line can be an edge (first difference) or bar (second difference) in an image.…”

Section: Facet Modellermentioning

confidence: 99%

An Interactive CAD-Based Vision System.

M-Rouhani¹,

Worrall²,

Anderson³

1995

Procedings of the British Machine Vision Conference 1995

View full text Add to dashboard Cite

We report a preliminary implementation of an interactive system that enables a draftsperson using only the standard features of a commercial CAD package (AutoCAD) to construct a model-based vision program. Some of the information needed by the model-based vision program is computed automatically from the drawing, but the draftsperson can also provides guidance on expected image-feature types by drawing the corresponding parts of the model on named CAD-layers and can indicate the visibility of arbitrary parts, such as expected shadow lines, by connecting them to modelled parts with invisible construction lines. Our system, running on a workstation, converts small drawings to computer vision programs in approximately 5 seconds, though an additional 35 seconds is used on networking, windowing, and software initialisation.We argue that interactive CAD-based vision compilers are more useful than contemporary batch-mode CAD-based vision compilers and conclude that the emergence of the international standard STEP will support the development of more competent CAD-based vision compilers of all kinds.

show abstract

“…It operates by evaluating evidence for straight lines in one or more images of an object [20]. A line can be an edge (first difference) or bar (second difference) in an image.…”

Section: Facet Modellermentioning

confidence: 99%

An Interactive CAD-Based Vision System.

M-Rouhani¹,

Worrall²,

Anderson³

1995

Procedings of the British Machine Vision Conference 1995

View full text Add to dashboard Cite

show abstract

“…We report experiments to demonstrate the performance of the modified algorithms. The output of the algorithms provides initial pose and class estimates which can be used to bootstrap model-based pose refinement (e.g., [35]) and vehicle tracking (see [6], [22], and [23]] for examples of tracking systems).…”

Section: Introductionmentioning

confidence: 99%

Efficient image gradient based vehicle localization

Tan

Baker²

2000

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

This paper reports novel algorithms for the efficient localization and recognition of traffic in traffic scenes. The algorithms eliminate the need for explicit symbolic feature extraction and matching. The pose and class of an object is determined by a form of voting and one-dimensional (1-D) correlations based directly on image gradient data, which can be computed "on the fly." The algorithms are therefore very well suited to real-time implementation. The algorithms make use of two a priori sources of knowledge about the scene and the objects expected: (1) the ground-plane constraint and (2) the fact that the overall shape of road vehicles is strongly rectilinear. Additional efficiency is derived from making the weak perspective assumption. These assumptions are valid in the road traffic application domain. The algorithms are demonstrated and tested using routine outdoor traffic images. Success with a variety of vehicles in several traffic scenes demonstrates the efficiency and robustness of context-based image understanding in road traffic scene analysis. The limitations of the algorithms are also addressed in the paper.

show abstract

“…The open world scene however can be a complex image to analyse, particularly due to illumination variations within the image and the changing pose of the object, each of which complicate the frame to frame matching of objects moving within the scene. Feature based geometric model matching, [2,4,10] has been shown to be very successful for identifying and tracking objects moving within a open world image, where the objects to be tracked occupy a significant proportion of the image; however they are less successful when the object to be tracked is further away from the camera and hence only occupying a small proportion of the image. In this case it has been found that the matching of crude object descriptors is more robust, [1,3].…”

Section: Introductionmentioning

confidence: 99%

Spatial-Temporal Reasoning Based on Object Motion

Teal

Ellis²

1996

Procedings of the British Machine Vision Conference 1996

View full text Add to dashboard Cite

This paper describes the continuing development of a system for tracking multiple man made objects, (typically vehicles) moving in a natural open world scene, where the detected motion is used to construct a structural representation of the scene. The system assumes no a priori knowledge of any structure within the image, but begins building a map of the scene on a frame by frame basis. The map shows regions in the image where vehicles are likely to be detected and regions where they are likely to become occluded. Tracking is complicated by the fact that the vehicles to be tracked are expected to be a large distance from the camera and as such will only occupy a small number of image pixels. The system has been tested using an input sequence of vehicles moving in a complex outdoor scene, where the vehicles undergo both full and partial occlusion.

show abstract

Advances in Model-Based Traffic Vision

Cited by 22 publications

References 7 publications

An Interactive CAD-Based Vision System.

An Interactive CAD-Based Vision System.

Efficient image gradient based vehicle localization

Spatial-Temporal Reasoning Based on Object Motion

Contact Info

Product

Resources

About