Vehicle Trajectory Reconstruction Research Articles

A longitudinal scanline based vehicle trajectory reconstruction method for high-angle traffic video

In this paper, a robust and efficient High-angle Spatial-Temporal Diagram Analysis (HASDA) model is built to reconstruct high-resolution vehicle trajectories from infrastructure traffic surveillance videos. A combined methodology was developed, comprising of scanline-based trajectory extraction and feature-matching coordinate transformation. A scanline-based trajectory extraction technique is introduced to separate vehicle strands from pavement background on the spatial-temporal diagram by considering color features, gradient features, and motion features. Particular cleaning algorithms for removing static object noises, shadows, and occlusions are also established. Feature-matching coordinate transformation converts the pixel coordinates to the real-world coordinates to generate the physical vehicle trajectory. To evaluate the algorithm, generated trajectory results were compared to the reconstructed version of the Next Generation Simulation (NGSIM) dataset. 15-min NGSIM video was divided into a 5-min dataset for the calibration and the remaining 10-min data for evaluation. Model parameters calibrated based on the 5-min video data are then applied to the 10-min testing data. Two levels of performance measurements are considered to evaluate both trajectory-level and point-level results. A reference algorithm based on mainstream motion-based detection and tracking methods are used as a baseline algorithm. Based on the evaluation results, the proposed method shows promising trajectory detection results, that on average more than 90% of vehicle trajectories are constructed by the proposed methods from the NGSIM videos. The HASDA model outperforms the reference algorithm and shows superior transferability in the training-testing experiment. Further work needs to be done to improve the algorithm performance against shadows and occlusions by incorporating more intelligent and advanced techniques.

Comparative Efficiency Analysis of Data Fusion Methods for Vehicle Trajectory Reconstruction

This article compares the efficiency of vehicle trajectory analysis methods based on data fusion from multiple cameras, monitoring the same area from different views under the condition having detection errors, which causes incorrectly localized and, in some cases, undetected vehicle during the movement. The experiment used the simulation of detection and localization of vehicle moving in straight, curved, zigzag and arbitrary trajectories, with localization errors and multi-level loss of data. By comparing Kalman-filter-based method and Linear-interpolation-based method for analyzing and reconstructing vehicle trajectory, the result shows that the data loss robustness of Kalman-filter-based method is higher than that of Linear-interpolation-based method, with data loss around 97% 97% and 90% for straight, curved and zigzag trajectories respectively. However, for arbitrary trajectory, the Linear-interpolation-based method is better than Kalman-filter-based method in all levels of data loss. In conclusion, Kalman-filter-based method is effective in the case of unchanged or slight transition of direction, while Linear-interpolation-based method is effective in the case of sudden transition of direction.

A generic data assimilation framework for vehicle trajectory reconstruction on signalized urban arterials using particle filters

With trajectory data, a complete microscopic and macroscopic picture of traffic flow operations can be obtained. However, trajectory data are difficult to observe over large spatiotemporal regions—particularly in urban contexts—due to practical, technical and financial constraints. The next best thing is to estimate plausible trajectories from whatever data are available. This paper presents a generic data assimilation framework to reconstruct such plausible trajectories on signalized urban arterials using microscopic traffic flow models and data from loops (individual vehicle passages and thus vehicle counts); traffic control data; and (sparse) travel time measurements from whatever source available. The key problem we address is that loops suffer from miss- and over-counts, which result in unbounded errors in vehicle accumulations, rendering trajectory reconstruction highly problematic. Our framework solves this problem in two ways. First, we correct the systematic error in vehicle accumulation by fusing the counts with sparsely available travel times. Second, the proposed framework uses particle filtering and an innovative hierarchical resampling scheme, which effectively integrates over the remaining error distribution, resulting in plausible trajectories. The proposed data assimilation framework is tested and validated using simulated data. Experiments and an extensive sensitivity analysis show that the proposed method is robust to errors both in the model and in the measurements, and provides good estimations for vehicle accumulation and vehicle trajectories with moderate sensor quality. The framework does not impose restrictions on the type of microscopic models used and can be naturally extended to include and estimate additional trajectory attributes such as destination and path, given data are available for assimilation.

Vehicle Collision Reconstruction With 3-D Inertial Navigation and GNSS

Complex and fatal road vehicle crashes are often a matter of dispute in forensic investigation. The proposed algorithms and methods, utilizing data from inertial measurement units and satellite navigation, enable road–vehicle trajectory reconstruction in 3-D suitable for postaccident forensic analysis. This paper provides an estimate of the reliability of this approach and its relation to component specifications. We start by describing an appropriate type of the strap-down inertial navigation system before providing a detailed performance analysis when the system is integrated with global navigation satellite systems. In addition, we present custom hardware and embedded software solutions, as well as experimental evaluation in real-world case study, capturing typical crash scenario performed at an approved safety performance assessment laboratory. In our experiments, strap-down inertial mechanization was used for short-term vehicle trajectory reconstruction using sensor readings and compared with ground-truth position captured on video. We have used reference equipment to validate the approach and the experiment resulted in a positioning accuracy on a decimeter level.

A new methodology for vehicle trajectory reconstruction based on wavelet analysis

Vehicle trajectories with high spatial and temporal resolution are known as the most ideal source of data for developing innovative microscopic traffic models. Aside from the method applied for collecting the vehicle trajectories, such data are more or less error-infected. The ever-increasing noise amplitude during the process of deriving the data (such as speed and acceleration) required for developing models, might change or even hide the structure of data and lead to useful information being overlooked. This highlights the importance of presenting the efficient methods which are adequate to remove noise and enhance the quality of vehicle trajectory data. Accordingly, in this paper a simple two-step technique based on wavelet analysis has been recommended for filtering errors and reconstructing trajectory data. Primarily, by using wavelet transform a special treatment was employed to identify and modify the outliers. Next, the noise in trajectory data was eliminated by applying the wavelet-based filter. The results of applying the proposed method to the synthetic noise-infected trajectory and the NGSIM dataset reveal how appropriate its performance is compared with other methodologies in terms of quantitative criteria.

Long queue estimation for signalized intersections using mobile data

Queue length is one of the key measures in assessing arterial performances. Under heavy congestion, queues are difficult to estimate from either fixed-location sensors (such as loop detectors) or mobile sensors since they may exceed the region of detection, which is defined as long queue in the literature. While the long queue problem has been successfully addressed in the past using fixed-location sensors, whether this can be done using mobile traffic sensors remains unclear. In this paper, a queue length estimation method is proposed to solve this long queue problem using short vehicle trajectories obtained from mobile sensors. The method contains vehicle trajectory reconstruction models to estimate the missing deceleration or acceleration process of a vehicle. Long queue estimation models are then developed using the reconstructed vehicle trajectories. The proposed method can provide estimates of the queue profile and the maximum queue length of a cycle. The method is tested in a field experiment with reasonable results.

Special issue: emerging technologies for intelligent transportation

Special issue: emerging technologies for intelligent transportation

A comparison of two Monte Carlo algorithms for 3D vehicle trajectory reconstruction in roundabouts

Visual vehicular trajectory analysis and reconstruction represent two relevant tasks both for safety and capacity concerns in road transportation. Especially in the presence of roundabouts, the perspective effects on vehicles projection on the image plane can be overcome by reconstructing their 3D positions with a 3D tracking algorithm. In this paper we compare two different Monte Carlo approaches to 3D model-based tracking: the Viterbi algorithm and the Particle Smoother. We tested the algorithms on a simulated dataset and on real data collected in one working roundabout with two different setups (single and multiple cameras). The Viterbi algorithm estimates the Maximum A-Posteriori solution from a sample-based state discretization, but, thanks to its continuous state representation, the Particle Smoother overcomes the Viterbi algorithm showing better performance and accuracy.

Queue Profile Estimation in Congested Urban Networks with Probe Data

AbstractQueues at signalized intersections are the main cause of traffic delays and travel time variability in urban networks. In this article, we propose a method to estimate queue profiles that are traffic shockwave polygons in the time‐space plane describing the spatiotemporal formation and dissipation of queues. The method integrates the collective effect of dispersed probe vehicle data with traffic flow shockwave analysis and data mining techniques. The proposed queue profile estimation method requires position and velocity data of probe vehicles; however, any explicit information of signal settings and arrival distribution is indispensable. Moreover, the method captures interdependencies in queue evolutions of successive intersections. The significance of the proposed method is that it is applicable in oversaturated conditions and includes queue spillover identification. Numerical results of simulation experiments and tests on NGSIM field data, with various penetration rates and sampling intervals, reveal the promising and robust performance of the proposed method compared with a uniform arrival queue estimation procedure. The method provides a thorough understanding of urban traffic flow dynamics and has direct applications for delay analysis, queue length estimation, signal settings estimation, and vehicle trajectory reconstruction.

Vehicle trajectory reconstruction using automatic vehicle identification and traffic count data

SummaryThe origin–destination (OD) matrix and the vehicle trajectory data are critical to transportation planning, design, and operation management. On the basis of the deployment of automatic vehicle identification (AVI) technology in urban traffic networks in China, this study proposed a vehicle trajectory reconstruction method for a large‐scale network by using AVI and traditional detector data. Particle filter theory was employed as the framework for this method that combines five spatial‐temporal trajectory correction factors (i.e., the path consistency, the AVI measurability criterion, the travel time consistency, the gravity flow model, and the path‐link flow matching model) to estimate the trajectory of a vehicle. The probabilities of the most likely trajectories are updated by the Bayesian method to approximate the ‘true’ trajectory. The traffic network in the Beijing Olympic Park was selected as the test bed and was simulated by using VISSIM to create a complete set of vehicle trajectories. The accuracy of the resulting trajectory reconstruction exceeds 90% when the AVI coverage is only 50%, assuming an AVI detection error of 5% for a closed network and an open network. The average relative error of a static OD matrix is 11.05%. Although the accuracy of reconstruction exceeds 80% when the AVI coverage is between 50% and 40%, the accuracy of a defective product‐OD matrix decreases rapidly. The proposed method yields high estimation accuracy for the full trajectories of individual vehicles and the OD matrix, which demonstrates significant potential for traffic‐related applications. Copyright © 2014 John Wiley & Sons, Ltd.

Vehicle trajectory reconstruction for signalized intersections using mobile traffic sensors

Mobile traffic sensors – those move with the flow they are monitoring – have quickly emerged as an important means for traffic monitoring and data collection. In this paper, we propose methods to reconstruct short vehicle trajectories for the entire traffic flow at arterial intersections using sample vehicle trajectories obtained from mobile traffic sensors. The motivation for this work is that once the trajectories for all vehicles are available, a complete picture of the traffic flow will be obtained. Such information can then be applied for arterial performance measurement and other related applications. This paper presents optimization-based and delay-based models to estimate shockwave boundaries at signalized intersections, which are used to reconstruct short vehicle trajectories by applying the variation formulation (VF) of traffic flow. The models are tested using the NGSIM data and micro-simulation data. The results indicate that the proposed models are not very sensitive to penetration rates. In general, the optimization-based model outperforms the delay-based model.

Monitoring of Human Driver Behavior by Vehicle Trajectory Reconstruction for Transportation Safety Management

In this paper, we proposed a method of monitoring human driver by reconstructing trajectories which transportation vehicle followed. For safety and management of logistic transportation, it is important to monitor the states of driving behavior through the whole course of path. Since many accidents occur due to the reckless driving of the driver every year, continuous monitoring of the status of commercial vehicles is needed for safety through the entire path from start point to the destination. To monitor the reckless driving, we tried to monitor the trajectory of the vehicle by using vehicle's lateral acceleration signal. Using the correlation between steering angle and lateral acceleration, we could find the relationship between steering angle and acceleration, and finally it is possible to estimate the global direction of vehicle maneuvering. We conducted experiments to find the history of vehicle position on the curved road using Kalman Filter, and classified steering wheel condition (over-steering, and under-steering). The method is applied to central safety management system for safety control of vehicles transporting toxic gases.

Performance Evaluation of Automotive Radars Using Carrier-Phase Differential GPS

We describe a centimeter-level ground-truth system based on GPS relative positioning. One of the goals of the system is for performance evaluation of automotive radars by creating objective test procedures for the radars. In support of this goal, an effective scheme is used to handle signal attenuation or blockage that leads to the changes of satellite visible map and cycle slip. Multiplatform GPS measurements and low-cost vehicle dynamic sensor data are integrated for vehicle trajectory reconstruction. Experimental results are presented to illustrate the system's actual centimeter-level accuracy and superior performance over Kalman smoothing. The effectiveness of the proposed ground-truth system is demonstrated by evaluating mono-pulse short-range radars whose accuracy is about submeter level.

Trajectory reconstruction with uncertainty estimation using mosaic registration

This paper addresses the problem of estimating the 3D trajectory and associated uncertainty of an underwater autonomous vehicle from a set of images of the seabed taken by an onboard camera. The presented algorithms resort to the use of video mosaics and build upon previous work on image registration and visual pose estimation. The pose estimation is accomplished in two steps. Firstly, a video mosaic is created automatically, covering a region of interest of the seabed. Then, after associating a 3D referential for the mosaic, the estimation of the camera position from a new view of the scene becomes possible. The main contribution of this paper lies on the assessment of the performance of the 3D pose algorithms. In order to do this, an image sequence with available ground-truth is used for precise error measuring. A first-order error propagation analysis is presented, relating the uncertainty in the location of the match points with the uncertainty in the pose parameters. The importance of predicting the estimate uncertainty is emphasized by the fact that it can be used for comparing algorithms and for the on-line monitoring of the vehicle trajectory reconstruction quality. Several iterative and non-iterative pose estimation methods are discussed, differing both on the criteria being minimized and on the required information about the camera intrinsic parameters. This information ranges from the full knowledge of the parameters, to the case where they are estimated using self-calibration from an image sequence under pure rotation. The implemented pose algorithms are compared for the accuracy and estimate covariance.