Specifications

The results of the video analysis demonstrate the ability of the system to provide data that can

potentially be useful for the understanding of driver behavior. The primary advantage of this

approach is the ease with which measurements can be made. The system requires only a single

video camera and a simple calibration procedure. Another advantage of the system is that a large

amount of data can be collected relative to the effort and expense involved.

There are a number of disadvantages of the video analysis approach developed in this research.

First, the analysis cannot guarantee that all vehicles passing through the intersection will be

tracked or that classification of all objects as a vehicle or pedestrian is perfect. The analysis is

also limited to the tracking of vehicles along a single lane. Many vehicles are not successfully

tracked across the entire frame due to the inability to match vehicles across occlusions. It is

likely that a multiple camera system would need to be utilized in order to reliably track all

vehicles passing through an intersection. A multiple camera approach would require additional

calibration procedures and increase the effort involved to configure the system. A final drawback

of digital video analysis is that the analysis itself can be very time consuming. It cannot be

conducted in real time. However, with additional development effort, the analyses could be made

more efficient.

6. TRAFFIC DATA ANALYSIS APPLICATION (TDAA) USAGE GUIDE

The TDAA software is built as a graphical user interface within the data analysis tool MatLab.

MatLab provides a rapid prototyping environment where development efforts can be focused on

the overall approach used to solve a problem rather than the details. MatLab provides simple,

straightforward functions for performing a range of image processing tasks. While the list of

available tools in MatLab is large, for this application it was at times necessary to call upon other

software to perform some tasks. The following is a list of minimum software requirements for

the TDAA:

Operating System: Windows XP (http://www.windows.com) or

Linux 2.6 (http://www.kernel.org)

Base System: MatLab Version R14 (

http://www.mathworks.com)

Image Extraction: FFMPEG (http://www.ffmpeg.org)

Video Extraction: OEM Software—Windows

dvgrab—Linux (

http://www.kinodv.org/)

Video Replay: Mplayer—Windows

(http://www.microsoft.com/windows/windowsmedia/default.mspx) or

playdv—Linux (http://www.kinodv.org/)

SIFT: SIFT—(http://www.cs.ubc.ca/~lowe/keypoints/)

TDAA is under continued development aimed at increasing the number of features and

capabilities. Described in this document are the features associated with the first release of this

software. The software is limited to processing video from the standard camera–intersection

configuration described in Section 3. The next release of the software will be targeted at

processing video from multiple cameras located at a single intersection. The analysis of video

from multiple cameras will enable the precise localization of the 3D position of a vehicle in an