Single- and Multi-Sensor Techniques to Improve Accuracy of Urban Vehicle Classification. Park, U., Silva, F., Hou, M., Heidemann, J., Guiliano, G., Wang, X., & Prashar, N. Technical Report ISI-TR-2006-614, USC/Information Sciences Institute, April, 2006. submitted for review![Single- and Multi-Sensor Techniques to Improve Accuracy of Urban Vehicle Classification [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper abstract bibtex Vehicle traffic sensors are an essential part of urban traffic management systems. However, humans are still widely used when short-term traffic classification is required, because current traffic data collection systems do not meet requirements for high accuracy, rapid deployment, and low cost. This paper explores the use of a sensor network for vehicle traffic classification. We accomplish low cost and easy deployment by using simple embedded computers and inductive loop sensors that can be taped directly to the roadway. We explore several techniques to improve accuracy both at individual sensors and by combining readings from multiple sensors. Traditional approaches typically assume high-speed traffic; we instead focus on errors that are induced at low and varying speeds, since we expect sensors to often be deployed at locations where vehicles will slow or stop. The contributions of this paper are first, a comprehensive evaluation of approaches to reduce error: from feature definition, combining readings from a single sensor, and combining results from multiple independent sensors. Second, we evaluate these approaches against both on-line and off-line human observations, demonstrating sensor accuracy better than or slightly worse than on-line human classification depending on the similarity of the categories, and nearly optimal for length-based classification.
@TechReport{Park06a,
author = "Unkyu Park and Fabio Silva and Mengzhao Hou and John Heidemann and Genevive Guiliano and Xi Wang and Nikhil Prashar",
title = "Single- and Multi-Sensor Techniques to Improve Accuracy of Urban Vehicle Classification",
institution = "USC/Information Sciences Institute",
year = 2006,
sortdate = "2006-04-01",
project = "ilense, surese",
jsubject = "sensornet_fusion",
number = "ISI-TR-2006-614",
note = "submitted for review",
month = apr,
location = "johnh: pafile",
keywords = "sensornet, traffic classification, sensor fusion",
url = "http://www.isi.edu/%7ejohnh/PAPERS/Park06a.html",
pdfurl = "http://www.isi.edu/%7ejohnh/PAPERS/Park06a.pdf",
otherurl = "ftp://ftp.isi.edu/isi-pubs/tr-614.pdf",
abstract = "
Vehicle traffic sensors are an essential part of urban traffic
management systems. However, humans are still widely used when
short-term traffic classification is required, because current traffic
data collection systems do not meet requirements for high accuracy,
rapid deployment, and low cost. This paper explores the use of a
sensor network for vehicle traffic classification. We accomplish low
cost and easy deployment by using simple embedded computers and
inductive loop sensors that can be taped directly to the roadway. We
explore several techniques to improve accuracy both at individual
sensors and by combining readings from multiple sensors. Traditional
approaches typically assume high-speed traffic; we instead focus on
errors that are induced at low and varying speeds, since we expect
sensors to often be deployed at locations where vehicles will slow or
stop. The contributions of this paper are first, a comprehensive
evaluation of approaches to reduce error: from feature definition,
combining readings from a single sensor, and combining results from
multiple independent sensors. Second, we evaluate these approaches
against both on-line and off-line human observations, demonstrating
sensor accuracy better than or slightly worse than on-line human
classification depending on the similarity of the categories, and
nearly optimal for length-based classification.
",
}
Downloads: 0
{"_id":"BEbDLudFPTvpehpzd","bibbaseid":"park-silva-hou-heidemann-guiliano-wang-prashar-singleandmultisensortechniquestoimproveaccuracyofurbanvehicleclassification-2006","author_short":["Park, U.","Silva, F.","Hou, M.","Heidemann, J.","Guiliano, G.","Wang, X.","Prashar, N."],"bibdata":{"bibtype":"techreport","type":"techreport","author":[{"firstnames":["Unkyu"],"propositions":[],"lastnames":["Park"],"suffixes":[]},{"firstnames":["Fabio"],"propositions":[],"lastnames":["Silva"],"suffixes":[]},{"firstnames":["Mengzhao"],"propositions":[],"lastnames":["Hou"],"suffixes":[]},{"firstnames":["John"],"propositions":[],"lastnames":["Heidemann"],"suffixes":[]},{"firstnames":["Genevive"],"propositions":[],"lastnames":["Guiliano"],"suffixes":[]},{"firstnames":["Xi"],"propositions":[],"lastnames":["Wang"],"suffixes":[]},{"firstnames":["Nikhil"],"propositions":[],"lastnames":["Prashar"],"suffixes":[]}],"title":"Single- and Multi-Sensor Techniques to Improve Accuracy of Urban Vehicle Classification","institution":"USC/Information Sciences Institute","year":"2006","sortdate":"2006-04-01","project":"ilense, surese","jsubject":"sensornet_fusion","number":"ISI-TR-2006-614","note":"submitted for review","month":"April","location":"johnh: pafile","keywords":"sensornet, traffic classification, sensor fusion","url":"http://www.isi.edu/%7ejohnh/PAPERS/Park06a.html","pdfurl":"http://www.isi.edu/%7ejohnh/PAPERS/Park06a.pdf","otherurl":"ftp://ftp.isi.edu/isi-pubs/tr-614.pdf","abstract":"Vehicle traffic sensors are an essential part of urban traffic management systems. However, humans are still widely used when short-term traffic classification is required, because current traffic data collection systems do not meet requirements for high accuracy, rapid deployment, and low cost. This paper explores the use of a sensor network for vehicle traffic classification. We accomplish low cost and easy deployment by using simple embedded computers and inductive loop sensors that can be taped directly to the roadway. We explore several techniques to improve accuracy both at individual sensors and by combining readings from multiple sensors. Traditional approaches typically assume high-speed traffic; we instead focus on errors that are induced at low and varying speeds, since we expect sensors to often be deployed at locations where vehicles will slow or stop. The contributions of this paper are first, a comprehensive evaluation of approaches to reduce error: from feature definition, combining readings from a single sensor, and combining results from multiple independent sensors. Second, we evaluate these approaches against both on-line and off-line human observations, demonstrating sensor accuracy better than or slightly worse than on-line human classification depending on the similarity of the categories, and nearly optimal for length-based classification. ","bibtex":"@TechReport{Park06a,\n\tauthor = \"Unkyu Park and Fabio Silva and Mengzhao Hou and John Heidemann and Genevive Guiliano and Xi Wang and Nikhil Prashar\",\n\ttitle = \t\"Single- and Multi-Sensor Techniques to Improve Accuracy of Urban Vehicle Classification\",\n\tinstitution = \t\"USC/Information Sciences Institute\",\n\tyear = \t\t2006,\n\tsortdate = \"2006-04-01\",\n\tproject = \"ilense, surese\",\n\tjsubject = \"sensornet_fusion\",\n\tnumber =\t\"ISI-TR-2006-614\",\n\tnote = \"submitted for review\",\n\tmonth =\t\tapr,\n\tlocation =\t\"johnh: pafile\",\n\tkeywords =\t\"sensornet, traffic classification, sensor fusion\",\n\turl =\t\t\"http://www.isi.edu/%7ejohnh/PAPERS/Park06a.html\",\n\tpdfurl =\t\"http://www.isi.edu/%7ejohnh/PAPERS/Park06a.pdf\",\n\totherurl =\t\"ftp://ftp.isi.edu/isi-pubs/tr-614.pdf\",\n\tabstract = \"\nVehicle traffic sensors are an essential part of urban traffic\nmanagement systems. However, humans are still widely used when\nshort-term traffic classification is required, because current traffic\ndata collection systems do not meet requirements for high accuracy,\nrapid deployment, and low cost. This paper explores the use of a\nsensor network for vehicle traffic classification. We accomplish low\ncost and easy deployment by using simple embedded computers and\ninductive loop sensors that can be taped directly to the roadway. We\nexplore several techniques to improve accuracy both at individual\nsensors and by combining readings from multiple sensors. Traditional\napproaches typically assume high-speed traffic; we instead focus on\nerrors that are induced at low and varying speeds, since we expect\nsensors to often be deployed at locations where vehicles will slow or\nstop. The contributions of this paper are first, a comprehensive\nevaluation of approaches to reduce error: from feature definition,\ncombining readings from a single sensor, and combining results from\nmultiple independent sensors. Second, we evaluate these approaches\nagainst both on-line and off-line human observations, demonstrating\nsensor accuracy better than or slightly worse than on-line human\nclassification depending on the similarity of the categories, and\nnearly optimal for length-based classification.\n\",\n}\n\n","author_short":["Park, U.","Silva, F.","Hou, M.","Heidemann, J.","Guiliano, G.","Wang, X.","Prashar, N."],"bibbaseid":"park-silva-hou-heidemann-guiliano-wang-prashar-singleandmultisensortechniquestoimproveaccuracyofurbanvehicleclassification-2006","role":"author","urls":{"Paper":"http://www.isi.edu/%7ejohnh/PAPERS/Park06a.html"},"keyword":["sensornet","traffic classification","sensor fusion"],"metadata":{"authorlinks":{}}},"bibtype":"techreport","biburl":"https://bibbase.org/f/dHevizJoWEhWowz8q/johnh-2023-2.bib","dataSources":["YLyu3mj3xsBeoqiHK","fLZcDgNSoSuatv6aX","fxEParwu2ZfurScPY","7nuQvtHTqKrLmgu99"],"keywords":["sensornet","traffic classification","sensor fusion"],"search_terms":["single","multi","sensor","techniques","improve","accuracy","urban","vehicle","classification","park","silva","hou","heidemann","guiliano","wang","prashar"],"title":"Single- and Multi-Sensor Techniques to Improve Accuracy of Urban Vehicle Classification","year":2006}