A novel image-based convolutional neural network approach for traffic congestion estimation
作者：
Highlights：
• A specific and accurate traffic congestion definition is proposed.
• Addressing the well-concerned issues on low estimation accuracy in bad conditions.
• A traffic congestion estimation method via Convolutional Neural Network is devised.
• Our Proposed Method estimates traffic congestion directly once the model trained.
• Traffic congestion estimation in free and congested conditions is evaluated.
摘要
•A specific and accurate traffic congestion definition is proposed.•Addressing the well-concerned issues on low estimation accuracy in bad conditions.•A traffic congestion estimation method via Convolutional Neural Network is devised.•Our Proposed Method estimates traffic congestion directly once the model trained.•Traffic congestion estimation in free and congested conditions is evaluated.
论文关键词：Traffic congestion,Convolutional neural network,Vehicle detection,Deep learning,Traffic flow parameter
论文评审过程：Received 19 November 2020, Revised 21 February 2021, Accepted 10 April 2021, Available online 2 May 2021, Version of Record 7 May 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.115037
