Share Email Print
cover

Proceedings Paper

4DCAPTURE: a general purpose software package for capturing and analyzing two-and three-dimensional motion data acquired from video sequences
Author(s): James S. Walton; Peter Hodgson; Karen Hallamasek; Jake Palmer
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

4DVideo is creating a general purpose capability for capturing and analyzing kinematic data from video sequences in near real-time. The core element of this capability is a software package designed for the PC platform. The software ("4DCapture") is designed to capture and manipulate customized AVI files that can contain a variety of synchronized data streams -- including audio, video, centroid locations -- and signals acquired from more traditional sources (such as accelerometers and strain gauges.) The code includes simultaneous capture or playback of multiple video streams, and linear editing of the images (together with the ancilliary data embedded in the files). Corresponding landmarks seen from two or more views are matched automatically, and photogrammetric algorithms permit multiple landmarks to be tracked in two- and three-dimensions -- with or without lens calibrations. Trajectory data can be processed within the main application or they can be exported to a spreadsheet where they can be processed or passed along to a more sophisticated, stand-alone, data analysis application. Previous attempts to develop such applications for high-speed imaging have been limited in their scope, or by the complexity of the application itself. 4DVideo has devised a friendly ("FlowStack") user interface that assists the end-user to capture and treat image sequences in a natural progression. 4DCapture employs the AVI 2.0 standard and DirectX technology which effectively eliminates the file size limitations found in older applications. In early tests, 4DVideo has streamed three RS-170 video sources to disk for more than an hour without loss of data. At this time, the software can acquire video sequences in three ways: (1) directly, from up to three hard-wired cameras supplying RS-170 (monochrome) signals; (2) directly, from a single camera or video recorder supplying an NTSC (color) signal; and (3) by importing existing video streams in the AVI 1.0 or AVI 2.0 formats. The latter is particularly useful for high-speed applications where the raw images are often captured and stored by the camera before being downloaded. Provision has been made to synchronize data acquired from any combination of these video sources using audio and visual "tags." Additional "front-ends," designed for digital cameras, are anticipated.

Paper Details

Date Published: 1 August 2003
PDF: 10 pages
Proc. SPIE 4948, 25th International Congress on High-Speed Photography and Photonics, (1 August 2003); doi: 10.1117/12.516775
Show Author Affiliations
James S. Walton, 4DVideo (United States)
Peter Hodgson, 4DVideo (United States)
Karen Hallamasek, 4DVideo (United States)
Jake Palmer, 4DVideo (United States)


Published in SPIE Proceedings Vol. 4948:
25th International Congress on High-Speed Photography and Photonics
Claude Cavailler; Graham P. Haddleton; Manfred Hugenschmidt, Editor(s)

© SPIE. Terms of Use
Back to Top