Vision-based vehicle detection and tracking method for forward collision warning in automobiles

Author

Srinivasa, Narayan

Author_Institution

HRL Labs. LLC, Malibu, CA, USA

Volume

2

fYear

2002

fDate

17-21 June 2002

Firstpage

626

Abstract

We describe a vision-based vehicle detection and tracking method for forward collision warning in automobiles. The approach is based on a set of edge-based constraint filters that assist in the segmentation of vehicles from background clutter. The detected vehicles are then tracked using a combination of distance based matching, sum-of-square-of-difference in intensity (SSD) and edge density of detected vehicle regions. The computational load for tracking is minimized using a vehicle-clustering algorithm. Experimental results are presented to illustrate the performance of the algorithm.

Keywords

automotive electronics; computational complexity; edge detection; image segmentation; optical tracking; SSD; automobiles; computational load minimization; distance-based matching; edge density; edge-based constraint filters; forward collision warning; image segmentation; intensity sum-of-square-of-difference; vehicle segmentation; vehicle-clustering algorithm; vision-based tracking; vision-based vehicle detection; Automobiles; Cameras; Image edge detection; Radar detection; Radar tracking; Road accidents; Sensor systems; Target tracking; Vehicle detection; Vehicles;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Vehicle Symposium, 2002. IEEE

Print_ISBN

0-7803-7346-4

Type

conf

DOI

10.1109/IVS.2002.1188021

Filename

1188021