TRB Pubsindex
Text Size:
x

« Prev Next »

Return to Results | New Search

Bookmark and Share

Title:

Video-Based Traffic Parameters Detecting at Intersection Entrance Using Foreground Spatio-Temporal Images

Accession Number:

01658358

Record Type:

Component

Abstract:

For efficient establishment of urban intelligent transportation system, accurate traffic parameters, such as the number of passed vehicles, occupancy and velocity, collected by surveillance video cameras installed at intersection entrance is critical. However, the serious vehicle occlusion, temporarily stopped or slow-moving vehicles and various road-user of urban intersection make it an extremely challenging task. In this paper, a novel algorithm for measurement of traffic parameters using foreground spatio-temporal images is presented to resolve deficiencies of traditional parameter collection methods, which may be highly computationally expensive or become unsuccessful with increasing complexity of traffic scenes. The algorithm is comprised of four steps: user initialization, building background model, horizontal and vertical foreground spatio-temporal images generation and foreground spatio-temporal images analysis. Traffic parameters rely on the horizontal and vertical foreground spatio-temporal images without camera calibration and vehicle tracking. The proposed algorithm was implemented in Visual C++ by OpenCV 3.3. Experimental results showed that the accuracy of vehicle counting is more then 90% in real-world urban traffic video database of diverse operating conditions such as daytime, shadowy daytime and nighttime, and it performs better than the state-of the-art methods, and real-time operation could be realized for slow-moving or high-speed vehicles.

Supplemental Notes:

This paper was sponsored by TRB committee ABJ50 Standing Committee on Information Systems and Technology.

Report/Paper Numbers:

18-02834

Language:

English

Authors:

Zhang, Yunsheng
Leng, Kaijun
Shi, Wen
Li, Hao

Pagination:

15p

Publication Date:

2018

Conference:

Transportation Research Board 97th Annual Meeting

Location: Washington DC, United States
Date: 2018-1-7 to 2018-1-11
Sponsors: Transportation Research Board

Media Type:

Digital/other

Features:

Figures; Photos; References; Tables

Uncontrolled Terms:

Temporal analysis

Subject Areas:

Data and Information Technology; Highways; Planning and Forecasting

Source Data:

Transportation Research Board Annual Meeting 2018 Paper #18-02834

Files:

TRIS, TRB, ATRI

Created Date:

Jan 8 2018 10:41AM