Shape Features of Overlapping Boundary for Classification of Moving Vehicles

Shape Features of Overlapping Boundary for Classification of Moving Vehicles

Elham Dallalzadeh (Department of Computer, Islamic Azad University Marvdasht Branch, Marvdasht, Iran) and D. S. Guru (Department of Studies in Computer Science, University of Mysore, Mysore, Karnataka, India)
Copyright: © 2013 |Pages: 13
DOI: 10.4018/ijcvip.2013010104
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

In this paper, the authors propose a model for classification of moving vehicles in traffic videos. A corner-based tracking method is presented to track and detect moving vehicles. The authors propose to overlap the boundary curves of each of the detected moving vehicles while tracking in a sequence of frames to reconstruct a complete boundary shape of the vehicle. The reconstructed boundary shape is normalized and then shape features are extracted. Vehicles are categorized into 4 different types of vehicle classes using KNN rule, the weighted KNN, PNN, and SVM classifiers. Experiments are conducted on traffic video sequences captured in an uncontrolled environment during daytime.
Article Preview

In literature, we can find a number of works on classification of vehicles in traffic videos. Sullivan et al., (1997) proposed a 3D model matching scheme to classify vehicles into various types like wagon, sedan, hatchback, etc. Although, 3D features obtained from stereo cameras might be useful for categorizing different classes of vehicles, the computational time of 3D model approaches are very high and classification of vehicles relies on detailed geometric of various types of traffic vehicles which might not be available all the time.

In Buch et al. (2009), they utilized a combined detector and classifier based on 3D wire frame models to locate ground plane position of vehicles. They generated motion silhouettes for an input video frame. The motion silhouettes are then applied to generate vehicle hypotheses. The classifier matches 3D wire frame models with the motion silhouettes. To validate the hypotheses, the normalized overlap area of motion silhouettes and projected model silhouettes are calculated. The classification results are used in the data association of the tracker to improve consistency and noise suppression.

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 7: 4 Issues (2017): 3 Released, 1 Forthcoming
Volume 6: 2 Issues (2016)
Volume 5: 2 Issues (2015)
Volume 4: 2 Issues (2014)
Volume 3: 4 Issues (2013)
Volume 2: 4 Issues (2012)
Volume 1: 4 Issues (2011)
View Complete Journal Contents Listing