Vehicle occlusion in congested ground traffic situations causes performance degradation in visual traffic surveillance systems. In this paper, we present a Hidden Markov Model (HMM) -based vehicle detection algorithm that is capable of handling vehicle occlusion and detecting vehicles from image sequences. In our algorithm, we first use Principal Component Analysis (PCA) and Multiple Discriminant Analysis (MDA) to extract features from input images, and then apply HMM to classify each image into three categories (road, head and body), where categories are called states in this paper. Finally we detect vehicles by analyzing the extracted state sequences. Results of experiments demonstrate that our algorithm is effective in congested traffic situations.
An HMM-Based Algorithm for Vehicle Detection in Congested Traffic Situations
2007-09-01
3788746 byte
Conference paper
Electronic Resource
English
Automated On-ramp Merging System for Congested Traffic Situations
Tema Archive | 2011
|Automated On-Ramp Merging System for Congested Traffic Situations
Online Contents | 2011
|Road traffic characteristics, driving patterns and emission factors for congested situations
TIBKAT | 2006
|