Urban Traffic Density Estimation Based on Ultrahigh-Resolution UAV Video and Deep Neural Network

Abstract

This paper presents an advanced urban traffic density estimation solution using the latest deep learning techniques to intelligently process ultra high resolution traffic videos taken from an Unmanned Aerial Vehicle (UAV). We first capture nearly an hour-long ultra high resolution traffic video at 5 busy road intersections of a modern megacity by flying an UAV during the rush hours. We then randomly sampled over 17K 512x512 pixel image patches from the video frames and manually annotated over 64K vehicles to form a dataset for this research which will also be made available to the research community for research purposes. Our innovative urban traffics analysis solution consists of advanced deep neural network based vehicle detection and localization, type (car, bus and truck) recognition, tracking and vehicle counting over time. We will present extensive experimental results to demonstrate the effectiveness of our solution. We will show that our enhanced Single Shot Multibox Detector (Enhanced-SSD) outperforms other deep neural network based techniques and that deep learning techniques are more effective than traditional computer vision techniques in traffic video analysis. We will also show that ultra high resolution video provides more information which enables more accurate vehicle detection and recognition than lower resolution contents. This paper not only demonstrates the advantages of using the latest technological advancements (ultra high resolution video and UAV) but also provides an advanced deep neural network based solution for exploiting these technological advancements for urban traffic density estimation.

Applications of unmanned aerial vehicle (UAV) in road safety, traffic and highway infrastructure management: Recent advances and challenges

“…Heat map produce for more robust detection Vehicle count, speed could be determined. Zhu et al (2018) Video data for calibration of model (manual annotated vehicles) Advanced deep Neural network model was trained for vehicle detection, localization, tracking and vehicle counting over time. Deep learning technologies are more effective than traditional vision-based algorithm.…”

Tracking and Simulating Pedestrian Movements at Intersections Using Unmanned Aerial Vehicles

“…To obtain high-quality data and reduce noisy information contained in traffic videos, we use a UAV to capture pedestrians, cyclists, and vehicles at road intersections. Compared with a traditional camera installed along the roadside or at intersections [45], the UAV surveillance approach has the following advantages: the hovering location and flying height of the UAV can be conveniently set and changed; the camera scope is substantially greater than that of traditional cameras; and high-resolution UAV videos can simultaneously capture richer information about pedestrians, cyclists, and vehicles.…”

“…Hence, counting was done by simply tallying the number of bounding boxes. We quantitatively evaluated the counting result via the correctness (Cor), completeness (Com), and quality (Qua), which are defined in [45] as:…”

“…A UAV is an effective tool to monitor geographical contexts with simple deployment and a low cost [43][44][45][46]. This study employs a UAV to automatically identify and track moving objects, including pedestrians, cyclists, and vehicles at complex traffic intersections, and to simulate the pedestrian movements.…”

CMRS: A Classifier Matrix Recognition System for Traffic Management and Analysis in a Smart City Environment

“…The processing features of this field inherit the advantages of IoT and other computing devices for augmenting the reliability of robust smart city solutions. Computer vision is employed in digital image processing, surveillance system actuation, road-traffic monitoring, medical image diagnosis and robotic systems [25]. The contributions of the paper are as follows.…”