Empirical Analysis of Vehicle Tracking Algorithms for Extracting Integral Trajectories from Consecutive Videos

IF 0.8 4区工程技术 Q4 TRANSPORTATION SCIENCE & TECHNOLOGY

Promet-Traffic & Transportation Pub Date : 2022-07-07 DOI:10.7307/ptt.v34i4.4041

Quan Chen, Hao Wang, Changyin Dong

{"title":"Empirical Analysis of Vehicle Tracking Algorithms for Extracting Integral Trajectories from Consecutive Videos","authors":"Quan Chen, Hao Wang, Changyin Dong","doi":"10.7307/ptt.v34i4.4041","DOIUrl":null,"url":null,"abstract":"This study introduces a novel methodological frame-work for extracting integral vehicle trajectories from several consecutive pictures automatically. The frame-work contains camera observation, eliminating image distortions, video stabilising, stitching images, identify-ing vehicles and tracking vehicles. Observation videos of four sections in South Fengtai Road, Nanjing, Jiangsu Province, China are taken as a case study to validate the framework. As key points, six typical tracking algorithms, including boosting, CSRT, KCF, median flow, MIL and MOSSE, are compared in terms of tracking reliability, operational time, random access memory (RAM) usage and data accuracy. Main impact factors taken into con-sideration involve vehicle colours, zebra lines, lane lines, lamps, guide boards and image stitching seams. Based on empirical analysis, it is found that MOSSE requires the least operational time and RAM usage, whereas CSRT presents the best tracking reliability. In addition, all tracking algorithms produce reliable vehicle trajecto-ry and speed data if vehicles are tracked steadily.","PeriodicalId":54546,"journal":{"name":"Promet-Traffic & Transportation","volume":null,"pages":null},"PeriodicalIF":0.8000,"publicationDate":"2022-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Promet-Traffic & Transportation","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.7307/ptt.v34i4.4041","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"TRANSPORTATION SCIENCE & TECHNOLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

This study introduces a novel methodological frame-work for extracting integral vehicle trajectories from several consecutive pictures automatically. The frame-work contains camera observation, eliminating image distortions, video stabilising, stitching images, identify-ing vehicles and tracking vehicles. Observation videos of four sections in South Fengtai Road, Nanjing, Jiangsu Province, China are taken as a case study to validate the framework. As key points, six typical tracking algorithms, including boosting, CSRT, KCF, median flow, MIL and MOSSE, are compared in terms of tracking reliability, operational time, random access memory (RAM) usage and data accuracy. Main impact factors taken into con-sideration involve vehicle colours, zebra lines, lane lines, lamps, guide boards and image stitching seams. Based on empirical analysis, it is found that MOSSE requires the least operational time and RAM usage, whereas CSRT presents the best tracking reliability. In addition, all tracking algorithms produce reliable vehicle trajecto-ry and speed data if vehicles are tracked steadily.

查看原文本刊更多论文

从连续视频中提取积分轨迹的车辆跟踪算法实证分析

本文提出了一种从连续图像中自动提取车辆整体轨迹的方法框架。该框架包括摄像头观察，消除图像失真，视频稳定，拼接图像，识别车辆和跟踪车辆。以中国江苏省南京市丰台南路四个路段的观测视频为例，对该框架进行了验证。重点比较了boost、CSRT、KCF、中位数流、MIL和MOSSE等六种典型跟踪算法的跟踪可靠性、运行时间、随机存取存储器(RAM)使用率和数据准确性。考虑的主要影响因素包括车辆颜色、斑马线、车道线、灯、导板和图像拼接接缝。通过实证分析，发现MOSSE对操作时间和内存占用最少，而CSRT具有最佳的跟踪可靠性。此外，如果车辆被稳定跟踪，所有跟踪算法都会产生可靠的车辆轨迹和速度数据。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Promet-Traffic & Transportation 工程技术-运输科技

CiteScore

1.90

自引率

20.00%

发文量

审稿时长

3 months

期刊介绍： This scientific journal publishes scientific papers in the area of technical sciences, field of transport and traffic technology. The basic guidelines of the journal, which support the mission - promotion of transport science, are: relevancy of published papers and reviewer competency, established identity in the print and publishing profile, as well as other formal and informal details. The journal organisation consists of the Editorial Board, Editors, Reviewer Selection Committee and the Scientific Advisory Committee. The received papers are subject to peer review in accordance with the recommendations for international scientific journals. The papers published in the journal are placed in sections which explain their focus in more detail. The sections are: transportation economy, information and communication technology, intelligent transport systems, human-transport interaction, intermodal transport, education in traffic and transport, traffic planning, traffic and environment (ecology), traffic on motorways, traffic in the cities, transport and sustainable development, traffic and space, traffic infrastructure, traffic policy, transport engineering, transport law, safety and security in traffic, transport logistics, transport technology, transport telematics, internal transport, traffic management, science in traffic and transport, traffic engineering, transport in emergency situations, swarm intelligence in transportation engineering. The Journal also publishes information not subject to review, and classified under the following headings: book and other reviews, symposia, conferences and exhibitions, scientific cooperation, anniversaries, portraits, bibliographies, publisher information, news, etc.