Reliable and Rapid Traffic Congestion Detection Approach Based on Deep Residual Learning and Motion Trajectories

Mohamed A Abdelwahab,Mohamed Abdel-Nasser,Maiya Hori

doi:10.1109/access.2020.3028395

Mohamed A Abdelwahab, Mohamed Abdel-Nasser + Show 1 more

Open Access

https://doi.org/10.1109/access.2020.3028395

Copy DOI

Abstract

Traffic congestion detection systems help manage traffic in crowded cities by analyzing videos of vehicles. Existing systems largely depend on texture and motion features. Such systems face several challenges, including illumination changes caused by variations in weather conditions, complexity of scenes, vehicle occlusion, and the ambiguity of stopped vehicles. To overcome these issues, this article proposes a rapid and reliable traffic congestion detection method based on the modeling of video dynamics using deep residual learning and motion trajectories. The proposed method efficiently uses both motion and deep texture features to overcome the limitations of existing methods. Unlike other methods that simply extract texture features from a single frame, we use an efficient representation learning method to capture the latent structures in traffic videos by modeling the evolution of texture features. This representation yields a noticeable improvement in detection results under various weather conditions. Regarding motion features, we propose an algorithm to distinguish stopped vehicles and background objects, whereas most existing motion-based approaches fail to address this issue. Both types of obtained features are used to construct an ensemble classification model based on the support vector machine algorithm. Two benchmark datasets are considered to demonstrate the robustness of the proposed method: the UCSD dataset and NU1 video dataset. The proposed method achieves competitive results (97.64% accuracy) when compared to state-of-the-art methods.

Highlights

Large cities contain millions of people that use different means of transportation, including buses, taxies, motorcycles, and bicycles
This study proposes a reliable and rapid traffic congestion classification method based on the modeling of video dynamics using a deep residual network and motion trajectories
We present a feasible method for modeling the dynamics of traffic videos based on representation learning and deep residual learning

Summary

Introduction

Large cities contain millions of people that use different means of transportation, including buses, taxies, motorcycles, and bicycles. The transportation infrastructure in large cities cannot accommodate continuous growth in the number of vehicles, which leads to traffic congestion. Additional sources of traffic congestion are weather conditions, traffic demand, and traffic-influencing events, such as accidents [1]. Stopping and starting vehicles in traffic gridlock consumes more fuel than normal traffic operations, leading to additional air pollution. Transport policy makers in several large cities have exploited intelligent traffic management systems (TMSs) to mitigate and prevent traffic congestion and improve overall traffic efficiency. TMSs include automated traffic monitoring systems that analyze images/videos captured by closed-circuit television cameras [2] to detect the status of traffic (e.g., light, medium, or heavy) and measure traffic flow

Objectives

Methods

Results

Conclusion