Object classification in traffic scenes using multiple spatio-temporal features

Guruprasad Somasundaram, Vassilios Morellas, Nikolaos P Papanikolopoulos, Saad Bedros

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

Object classification is a widely researched area in the field of computer vision. Lately there has been a lot of attention to appearance based models for representing objects. The most important feature of classifying objects such as pedestrians, vehicles, etc. in traffic scenes is that we have motion information available to us. The motion information presents itself in the form of temporal cues such as velocity and also as spatio-temporal cues such as optical flow, DHOG [6], etc. We propose a novel spatio-temporal feature based on covariance descriptors known as DCOV which captures complementary information to the DHOG feature. We present a combined classifier based on properties of tracked objects along with the DHOG and the DCOV features. We show based on experiments on the PETS 2001 dataset and two video sequences of bicycle and pedestrian traffic that the proposed classifier provides competent performance for distinguishing pedestrians, vehicles and bicyclists. Our method is also adaptive and benefits from the availability of more data for training. The classifier is also developed with real-time feasibility in mind.

Original languageEnglish (US)
Title of host publication2012 20th Mediterranean Conference on Control and Automation, MED 2012 - Conference Proceedings
Pages1536-1541
Number of pages6
DOIs
StatePublished - Oct 5 2012
Event2012 20th Mediterranean Conference on Control and Automation, MED 2012 - Barcelona, Spain
Duration: Jul 3 2012Jul 6 2012

Other

Other2012 20th Mediterranean Conference on Control and Automation, MED 2012
CountrySpain
CityBarcelona
Period7/3/127/6/12

Fingerprint

Classifiers
Bicycles
Optical flows
Computer vision
Availability
Experiments

Cite this

Somasundaram, G., Morellas, V., Papanikolopoulos, N. P., & Bedros, S. (2012). Object classification in traffic scenes using multiple spatio-temporal features. In 2012 20th Mediterranean Conference on Control and Automation, MED 2012 - Conference Proceedings (pp. 1536-1541). [6265857] https://doi.org/10.1109/MED.2012.6265857

Object classification in traffic scenes using multiple spatio-temporal features. / Somasundaram, Guruprasad; Morellas, Vassilios; Papanikolopoulos, Nikolaos P; Bedros, Saad.

2012 20th Mediterranean Conference on Control and Automation, MED 2012 - Conference Proceedings. 2012. p. 1536-1541 6265857.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Somasundaram, G, Morellas, V, Papanikolopoulos, NP & Bedros, S 2012, Object classification in traffic scenes using multiple spatio-temporal features. in 2012 20th Mediterranean Conference on Control and Automation, MED 2012 - Conference Proceedings., 6265857, pp. 1536-1541, 2012 20th Mediterranean Conference on Control and Automation, MED 2012, Barcelona, Spain, 7/3/12. https://doi.org/10.1109/MED.2012.6265857
Somasundaram G, Morellas V, Papanikolopoulos NP, Bedros S. Object classification in traffic scenes using multiple spatio-temporal features. In 2012 20th Mediterranean Conference on Control and Automation, MED 2012 - Conference Proceedings. 2012. p. 1536-1541. 6265857 https://doi.org/10.1109/MED.2012.6265857
Somasundaram, Guruprasad ; Morellas, Vassilios ; Papanikolopoulos, Nikolaos P ; Bedros, Saad. / Object classification in traffic scenes using multiple spatio-temporal features. 2012 20th Mediterranean Conference on Control and Automation, MED 2012 - Conference Proceedings. 2012. pp. 1536-1541
@inproceedings{e5d0b266c6f94e8d9aa95f3fabf88d92,
title = "Object classification in traffic scenes using multiple spatio-temporal features",
abstract = "Object classification is a widely researched area in the field of computer vision. Lately there has been a lot of attention to appearance based models for representing objects. The most important feature of classifying objects such as pedestrians, vehicles, etc. in traffic scenes is that we have motion information available to us. The motion information presents itself in the form of temporal cues such as velocity and also as spatio-temporal cues such as optical flow, DHOG [6], etc. We propose a novel spatio-temporal feature based on covariance descriptors known as DCOV which captures complementary information to the DHOG feature. We present a combined classifier based on properties of tracked objects along with the DHOG and the DCOV features. We show based on experiments on the PETS 2001 dataset and two video sequences of bicycle and pedestrian traffic that the proposed classifier provides competent performance for distinguishing pedestrians, vehicles and bicyclists. Our method is also adaptive and benefits from the availability of more data for training. The classifier is also developed with real-time feasibility in mind.",
author = "Guruprasad Somasundaram and Vassilios Morellas and Papanikolopoulos, {Nikolaos P} and Saad Bedros",
year = "2012",
month = "10",
day = "5",
doi = "10.1109/MED.2012.6265857",
language = "English (US)",
isbn = "9781467325318",
pages = "1536--1541",
booktitle = "2012 20th Mediterranean Conference on Control and Automation, MED 2012 - Conference Proceedings",

}

TY - GEN

T1 - Object classification in traffic scenes using multiple spatio-temporal features

AU - Somasundaram, Guruprasad

AU - Morellas, Vassilios

AU - Papanikolopoulos, Nikolaos P

AU - Bedros, Saad

PY - 2012/10/5

Y1 - 2012/10/5

N2 - Object classification is a widely researched area in the field of computer vision. Lately there has been a lot of attention to appearance based models for representing objects. The most important feature of classifying objects such as pedestrians, vehicles, etc. in traffic scenes is that we have motion information available to us. The motion information presents itself in the form of temporal cues such as velocity and also as spatio-temporal cues such as optical flow, DHOG [6], etc. We propose a novel spatio-temporal feature based on covariance descriptors known as DCOV which captures complementary information to the DHOG feature. We present a combined classifier based on properties of tracked objects along with the DHOG and the DCOV features. We show based on experiments on the PETS 2001 dataset and two video sequences of bicycle and pedestrian traffic that the proposed classifier provides competent performance for distinguishing pedestrians, vehicles and bicyclists. Our method is also adaptive and benefits from the availability of more data for training. The classifier is also developed with real-time feasibility in mind.

AB - Object classification is a widely researched area in the field of computer vision. Lately there has been a lot of attention to appearance based models for representing objects. The most important feature of classifying objects such as pedestrians, vehicles, etc. in traffic scenes is that we have motion information available to us. The motion information presents itself in the form of temporal cues such as velocity and also as spatio-temporal cues such as optical flow, DHOG [6], etc. We propose a novel spatio-temporal feature based on covariance descriptors known as DCOV which captures complementary information to the DHOG feature. We present a combined classifier based on properties of tracked objects along with the DHOG and the DCOV features. We show based on experiments on the PETS 2001 dataset and two video sequences of bicycle and pedestrian traffic that the proposed classifier provides competent performance for distinguishing pedestrians, vehicles and bicyclists. Our method is also adaptive and benefits from the availability of more data for training. The classifier is also developed with real-time feasibility in mind.

UR - http://www.scopus.com/inward/record.url?scp=84866910982&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84866910982&partnerID=8YFLogxK

U2 - 10.1109/MED.2012.6265857

DO - 10.1109/MED.2012.6265857

M3 - Conference contribution

SN - 9781467325318

SP - 1536

EP - 1541

BT - 2012 20th Mediterranean Conference on Control and Automation, MED 2012 - Conference Proceedings

ER -