Detection Of Vulnerable Road Users Research Articles

Sensor-Fused Nighttime System for Enhanced Pedestrian Detection in ADAS and Autonomous Vehicles.

Ensuring a safe nighttime environmental perception system relies on the early detection of vulnerable road users with minimal delay and high precision. This paper presents a sensor-fused nighttime environmental perception system by integrating data from thermal and RGB cameras. A new alignment algorithm is proposed to fuse the data from the two camera sensors. The proposed alignment procedure is crucial for effective sensor fusion. To develop a robust Deep Neural Network (DNN) system, nighttime thermal and RGB images were collected under various scenarios, creating a labeled dataset of 32,000 image pairs. Three fusion techniques were explored using transfer learning, alongside two single-sensor models using only RGB or thermal data. Five DNN models were developed and evaluated, with experimental results showing superior performance of fused models over non-fusion counterparts. The late-fusion system was selected for its optimal balance of accuracy and response time. For real-time inferencing, the best model was further optimized, achieving 33 fps on the embedded edge computing device, an 83.33% improvement in inference speed over the system without optimization. These findings are valuable for advancing Advanced Driver Assistance Systems (ADASs) and autonomous vehicle technologies, enhancing pedestrian detection during nighttime to improve road safety and reduce accidents.

On the Detection of Vulnerable Road Users—Simulating Limits of an FMCW Radar

The complexity of urban traffic environments poses unique challenges for automotive radar sensors. Limited range and Doppler resolution hinder the ability to distinguish closely spaced objects and detect vulnerable road users, particularly pedestrians. Therefore, it is important to know the limits of FMCW radars. To assess these limitations, the WinProp channel simulator is used to model and simulate traffic scenarios regarding the detection of the range, velocity, and SNR of targets. To validate WinProp as a simulation tool for traffic scenarios, a real traffic scenario is configured and the simulation results compared to the measurement results in order to demonstrate the suitability of WinProp for simulating traffic scenarios and evaluating radar detection performance. The close agreement between the simulation and measurement results provides validation for WinProp’s accuracy in modeling detections and signal processing within traffic scenarios. To showcase the limitations of FMCW radars, two scenarios with challenging detection conditions are simulated: an occluded pedestrian between parked vehicles and a densely populated road. A monostatic setup with a single patch antenna and a 16 × 16 patch antenna array is evaluated, using chirp bandwidths of 1 GHz and 4 GHz. The simulations have shown that although it is feasible to detect an occluded pedestrian by multipath propagation, there is still a need in improving the detection performance when there is no multipath from an occluded pedestrian. Furthermore, the results indicate that higher bandwidth improves target separation but is insufficient for occluded pedestrian detection without multipath. However, with some targets, even the range resolution at 4 GHz was not sufficient, which required a separation in the Doppler dimension, emphasizing the need for overall resolution improvement of FMCW radars. Future work should focus on developing more complex road user models and simulating a broader range of scenarios to comprehensively evaluate radar performance in road traffic environments.

Modeling and evaluation of cooperative Vulnerable Road User protection schemes in realistic C-ITS environments

Vulnerable Road Users (VRUs) comprising pedestrians, cyclists, and motorcyclists are more susceptible to road traffic accidents and fatalities than other road users. Improving safety for vulnerable road users is a priority for many cities and governments, and various measures are being taken to reduce the risk of accidents and increase road safety. Advanced Cooperative Intelligent Transport Systems (C-ITS) that facilitate Vehicle-to-Everything (V2X) communication encompass Vehicle-to-Pedestrian (V2P) schemes. To address the issues and decrease the number of VRU fatalities, efficient VRU protection solutions can be seamlessly incorporated into evolving C-ITS services. This paper introduces a novel implementation and integration of VRU protection models into the Artery/OMNeT++ V2X-simulation framework as a promising solution for further wide-scale evolution in enhancing the safety of vulnerable road users. VRUs can transmit messages about their existence using the VRU Awareness Message (VAM) messaging protocol. This paper presents a performance evaluation of cooperative VRU protection scenarios with VAM protocol and deployment. Simulation results will be given for pedestrian protection models with VRU detection using sensors, collective perception, and active, VAM-based VRU Awareness message transmission.

An Enhanced Detector for Vulnerable Road Users Using Infrastructure-Sensors-Enabled Device.

The precise and real-time detection of vulnerable road users (VRUs) using infrastructure-sensors-enabled devices is crucial for the advancement of intelligent traffic monitoring systems. To overcome the prevalent inefficiencies in VRU detection, this paper introduces an enhanced detector that utilizes a lightweight backbone network integrated with a parameterless attention mechanism. This integration significantly enhances the feature extraction capability for small targets within high-resolution images. Additionally, the design features a streamlined 'neck' and a dynamic detection head, both augmented with a pruning algorithm to reduce the model's parameter count and ensure a compact architecture. In collaboration with the specialized engineering dataset De_VRU, the model was deployed on the Hisilicon_Hi3516DV300 platform, specifically designed for infrastructure units. Rigorous ablation studies, employing YOLOv7-tiny as the baseline, confirm the detector's efficacy on the BDD100K and LLVIP datasets. The model not only achieved an improvement of over 12% in the mAP@50 metric but also realized a reduction in parameter count by more than 40%, and a 50% decrease in inference time. Visualization outcomes and a case study illustrate the detector's proficiency in conducting real-time detection with high-resolution imagery, underscoring its practical applicability.

An analysis of day and night bicyclist injury severities in vehicle/bicycle crashes: A comparison of unconstrained and partially constrained temporal modeling approaches

Due to visibility limitations and other factors, the injuries sustained by bicyclists in nighttime vehicle-bicycle crashes tend to be more severe than daytime crashes. This paper seeks to provide insights into this day/night injury severity phenomenon by studying how day/night bicyclist injury severities have changed in crashes that occurred before, during, and after the COVID-19 lock downs. Using data from vehicle-bicycle crashes in the state of Florida over a three-year period (from 2019 to 2021 inclusive), separate yearly models of bicyclist-injury severities (with possible outcomes of severe injury, minor injury, and no visible injury) were estimated using a random parameters logit approach with possible heterogeneity in the means and variances of random parameters. Likelihood ratio tests were conducted to examine the overall stability of model estimates across the studied years as well as day/night differences, and a comparison of partially constrained and unconstrained temporal modeling approaches was undertaken. A wide range of variables potentially affecting resulting bicyclist injury severities in vehicle/bicycle crashes was considered including bicyclist and vehicle driver information, vehicle features, roadways and environmental conditions, temporal characteristics, and roadway features. The findings show statistically significant injury-severity differences between daytime and nighttime before, during and after the COVID-19 pandemic. Out-of-sample simulation results suggest that improving the visibility of bicyclist through mandated reflectivity, improved roadway illumination, undertaking public awareness campaigns relating to nighttime bicyclist safety, and vulnerable road user detection sensors in vehicles can all contribute to substantially improving nighttime bicyclist safety.

An Improved YOLOv5 Algorithm for Vulnerable Road User Detection.

The vulnerable road users (VRUs), being small and exhibiting random movements, increase the difficulty of object detection of the autonomous emergency braking system for vulnerable road users AEBS-VRUs, with their behaviors highly random. To overcome existing problems of AEBS-VRU object detection, an enhanced YOLOv5 algorithm is proposed. While the Complete Intersection over Union-Loss (CIoU-Loss) and Distance Intersection over Union-Non-Maximum Suppression (DIoU-NMS) are fused to improve the model's convergent speed, the algorithm also incorporates a minor object detection layer to increase the performance of VRU detection. A dataset for complex AEBS-VRUS scenarios is established based on existing datasets such as Caltech, nuScenes, and Penn-Fudan, and the model is trained using migration learning based on the PyTorch framework. A number of comparative experiments using models such as YOLOv6, YOLOv7, YOLOv8 and YOLOx are carried out. The results of the comparative evaluation show that the proposed improved YOLO5 algorithm has the best overall performance in terms of efficiency, accuracy and timeliness of target detection.

Replacing the human driver: An objective benchmark for occluded pedestrian detection

Early detection of vulnerable road users is a crucial requirement for autonomous vehicles to meet and exceed the object detection capabilities of human drivers. One of the most complex outstanding challenges is that of partial occlusion where a target object is only partially available to the sensor due to obstruction by another foreground object. A number of leading pedestrian detection benchmarks provide annotation for partial occlusion, however each benchmark varies greatly in their definition of the occurrence and severity of occlusion. Research demonstrates that a high degree of subjectivity is used to classify occlusion level in these cases and occlusion is typically categorized into 2–3 broad categories such as “partially” and “heavily” occluded. In addition, many pedestrian instances are impacted by multiple inhibiting factors which contribute to non-detection such as object scale, distance from camera, lighting variations and adverse weather. This can lead to inaccurate or inconsistent reporting of detection performance for partially occluded pedestrians depending on which benchmark is used. This research introduces a novel, objective benchmark for partially occluded pedestrian detection to facilitate the objective characterization of pedestrian detection models. Characterization is carried out on seven popular pedestrian detection models for a range of occlusion levels from 0%–99% to demonstrate the impact of progressive levels of partial occlusion on pedestrian detectability. Results show that the proposed benchmark provides more objective, fine grained analysis of pedestrian detection algorithms than the current state of the art.

Improving 3D Vulnerable Road User Detection With Point Augmentation

Point clouds have been a popular representation to describe 3D environments for autonomous driving applications. Despite accurate depth information, sparsity of pointsresults in difficulties in extracting sufficient features from vulnerable objects of small sizes. One solution is leveraging self-attention networks to build long-range connections between similar objects. Another method is using generative models to estimate the complete shape of objects. Both approaches introduce large memory consumption and extra complexity to the models while the geometric characteristics of objects are overlooked. To overcome this problem, this paper proposes Point Augmentation (PA)- RCNN, focusing on small object detection by generating efficient complementary features without trainable parameters. Specifically, 3D points are sampled with the guidance of object proposals and encoded through the 3D grid-based feature aggregation to produce localised 3D voxel properties. Such voxel attributes are fed to the pooling module with the aid of fictional points, which are transformed from sampled points considering geometric symmetry. Experimental results on Waymo Open Dataset and KITTI dataset show a superior advantage in the detection of distant and small objects in comparison with existing state-ofthe- art methods.

Smart Spread Spectrum Modulated Tags for Detection of Vulnerable Road Users with Automotive Radar.

In recent years, there has been a significant increase in the number of collisions between vehicles and vulnerable road users such as pedestrians, cyclists, road workers and more recently scooter riders, especially in urban streets. This work studies the feasibility of enhancing the detection of these users by means of CW radars because they have a low radar cross section. Since the speed of these users is usually low, they can be confused with clutter due to the presence of large objects. To this end, this paper proposes, for the first time, a method based on a spread spectrum radio communication between vulnerable road users and the automotive radar consisting of modulating a backscatter tag, placed on the user. In addition, it is compatible with low-cost radars that use different waveforms such as CW, FSK or FMCW, and hardware modifications are not required. The prototype that has been developed is based on a commercial monolithic microwave integrated circuit (MMIC) amplifier connected between two antennas, which is modulated by switching its bias. Experimental results with a scooter, under static and moving conditions, using a low-power Doppler radar at a 24 GHz band compatible with blind spot radars, are provided.

The detection of vulnerable road users by younger and older drivers

The detection of vulnerable road users (VRUs), especially under time constraints, may be impaired in elderly drivers, due to their visual and cognitive decline. This represents a major concern for road safety. The objective of the present study was to investigate the effect of aging on the detection of VRUs. A further aim was to investigate the impact of external factors on VRU detection. Twenty-two young adults, 20 younger-old adults, and 32 older-old adults were included in the study. A series of photographs were displayed for 500 ms. Participants were asked to detect the VRU, which could be a motorcyclist, a cyclist, or a pedestrian. The VRU was located at one of two distances (near/far), one of two locations (off-centered/centered), and in two car-DRL (daytime running lights) environments (on/off). The ability to correctly detect a VRU was measured. An eye-tracker was used to record eye movements. The main findings showed that VRU-detection performance decreased with aging, even in the younger-old group (ages 55–68). The ability to correctly detect a VRU by older-old adults was poorer particularly when the VRU was a motorcyclist or a cyclist. As a whole, the older-old adults made more fixations to correctly detect the VRU than the other two age groups did. Moreover, the visual angle between the gaze of the participant’s last fixation and the target in the older-old group was lower than in the other two groups, particularly when the VRU was off-centered. This finding suggests that older-old adults compensated for their visual-field decline by doing more visual exploration than the other two groups did to correctly detect the VRU. The results are discussed with regards to age-related cognitive and visual deficits.Word count: 279.

A novel early warning strategy for right-turning blind zone based on vulnerable road users detection

A novel early warning strategy for right-turning blind zone based on vulnerable road users detection

On the safety of vulnerable road users by cyclist detection and tracking

Timely detection of vulnerable road users is of great relevance to avoid accidents in the context of intelligent transportation systems. In this work, detection and tracking is acknowledged for a particularly vulnerable class of road users, the cyclists. We present a performance comparison between the main deep learning-based algorithms reported in the literature for object detection, such as SSD, Faster R-CNN and R-FCN along with InceptionV2, ResNet50, ResNet101, Mobilenet V2 feature extractors. In order to identify the cyclist heading and predict its intentions, we propose a multi-class detection with eight classes according to orientations. To do so, we introduce a new dataset called “CIMAT-Cyclist”, containing 20,229 cyclist instances over 11,103 images, labeled based on the cyclist’s orientation. To improve the performance in cyclists’ detection, the Kalman filter is used for tracking, coupled together with the Kuhn–Munkres algorithm for multi-target association. Finally, the vulnerability of the cyclists is evaluated for each instance in the field of view, taking into account their proximity and predicted intentions according to their heading angle, and a risk level is assigned to each cyclist. Experimental results validate the proposed strategy in real scenarios, showing good performance.

Impact of Safety Message Generation Rules on the Awareness of Vulnerable Road Users.

In the face of cooperative intelligent transportation systems (C-ITS) advancements, the inclusion of vulnerable road users (VRU), i.e., pedestrians, cyclists, and motorcyclists, has just recently become a part of the discussion. Including VRU in C-ITS presents new challenges, most notably the trade-off between the increase in VRU safety and the aggravation in channel congestion resulting from VRU-generated messages. However, previous studies mainly focus on network-related metrics without giving much consideration to VRU safety-related metrics. In this context, we evaluated such a trade-off with a study of motion-based message generation rules for VRU transmissions. The rules were analyzed using theoretical and simulation-based evaluations. In addition to studying the message generation rules using channel load metrics, such as channel busy ratio (CBR) and packet delivery ratio (PDR), we introduced a new metric: the VRU Awareness Probability (VAP). VAP uses the exchange of messages from active VRU to measure the probability of VRU detection by nearby vehicles. Results show that fixed message-filtering mechanisms reduce the overall channel load, but they could negatively impact VRU detection. We established the importance of quantifying the VRU awareness and its inclusion in C-ITS analysis because of its direct impact on VRU safety. We also discussed approaches that include VRU context and dynamism to improve the definition of message generation rules.

Overcoming Occlusion in the Automotive Environment—A Review

Accurate and consistent vulnerable road user detection remains one of the most challenging perception tasks for autonomous vehicles. One of the most complex outstanding issues is partial occlusion, where a sensor has only a partial view of the target object due to a foreground object that partially obscures the target. A review of occlusion detection and handling solutions for the automotive environment is presented by this research. This article first discusses object detection by the human visual system, provides an overview of occlusion reasoning in computer vision, presents a summary of occlusion handling strategies in pedestrian, vehicle and object detection applications in the automotive environment. A selection of the remaining challenges to achieving the required level of object detection performance for safe autonomous driving are also discussed.

Intentions of Vulnerable Road Users—Detection and Forecasting by Means of Machine Learning

Avoiding collisions with vulnerable road users (VRUs) using sensor-based early recognition of critical situations is one of the manifold opportunities provided by the current development in the field of intelligent vehicles. As, especially, pedestrians and cyclists are very agile and have a variety of movement options, modeling their behavior in traffic scenes becomes a challenging task. In this paper, we propose movement models based on machine learning methods, in particular, artificial neural networks, in order to classify the current motion state and to predict the future trajectory of the VRUs. Both model types are also combined to enable the application of specifically trained motion predictors based on a continuously updated pseudo probabilistic state classification. Furthermore, the architecture is used to evaluate motion-specific physical models for starting and stopping and video-based pedestrian motion classification. A comprehensive dataset consisting of a total of 1068 pedestrian and 494 cyclist scenes acquired at an urban intersection is used for optimization, training, and evaluation of the different models. The results show substantially higher classification rates and the ability, through the machine learning approaches, to earlier recognize motion state changes than by the way of interacting multiple model (IMM) Kalman filtering. The trajectory prediction quality has also been improved for all kinds of test scenes, especially when starting and stopping motions are included. Here, 37% and 41% fewer position errors were achieved on average, respectively.

Does a yellow jacket enhance cyclists’ sensory conspicuity for car drivers during daylight hours in an urban environment?

The aim of this study was to find out if the colour contrast of a vest worn by cyclists improved their visibility for motorists in a simulated car driving task. BackgroundIn several cases involving collisions between cars and cyclists, car drivers failed to detect the latter in time to avoid collision because of their low visibility or sensory conspicuity. We studied the effect of a yellow cyclist jacket on the detection of cyclists by motorists in an urban environment during daylight hours. Method43 motorists performed a vulnerable road user detection task in a car-driving simulator. They had to detect cyclists and pedestrians in an urban setting, to locate them and estimate how difficult it was to detect them. Cyclists’ levels of visibility for car drivers were either high or low in critical situations. The cyclists were dressed in grey and wore yellow jackets, or they did not. ResultsMotorists detected cyclists wearing jackets at a greater distance, but only in those situations previously identified as being of high cyclist visibility. Several mechanisms underlying the sensory conspicuity of cyclists for car drivers were considered. Motorists were able to estimate cyclist visibility correctly based on the type of situation, but were unable to evaluate the jacket’s effect on cyclist visibility. ConclusionThe attentional selection of a cyclist in the road environment during car driving depends partly on bottom-up processing (such as saliency related to colour contrast). However, the yellow cyclist jacket proved to be insufficient as a visibility aid.

Driver’s emotional state and detection of vulnerable road users: Towards a better understanding of how emotions affect drivers’ perception using cardiac and ocular metrics

Traditionally, anger has been considered to have a detrimental effect on driving. However, recent studies suggest that this statement should be qualified, especially where vulnerable road user detection is concerned. One primary factor which may determine anger’s effect on a driver’s attention is its intensity. In the present study, different intensities of anger were elicited via film clips, then performances in vulnerable road user detection (i.e., of cyclists) were assessed while participants drove a car in a simulated environment. Cardiac and ocular measurements and self-reported data were used in order to accurately assess emotional state and attention management throughout the experiment. Results suggested that participants resorted to reappraisal strategies when they were exposed to the emotional film clips. This phenomenon did not directly affect cyclist detection performances, but evidence of different visual scanning strategies between groups emerged. The contribution of cardiac and ocular measurements to emotional assessment and the advantages of appraisal approaches of emotion were also discussed.

Mechanisms underlying cognitive conspicuity in the detection of cyclists by car drivers

ObjectiveThe aim of this study was to evaluate the visibility of cyclists for motorists in a simulated car driving task. BackgroundIn several cases involving collisions between cars and cyclists, car drivers failed to detect the latter in time to avoid collision because of their low conspicuity. Method2 groups of motorists (29.2 years old), including 12 cyclist-motorists and 13 non-cyclist-motorists, performed a vulnerable road user detection task in a car-driving simulator. They had to detect cyclists and pedestrians in an urban setting and evaluate the realism of the cyclists, the traffic, the city, the infrastructure, the car driven and the situations. Cyclists appeared in critical situations derived from previous accounts given by injured cyclists and from cyclists’ observations in real-life situations. Cyclist’s levels of visibility for car drivers were either high or low in these situations according to the cyclists. ResultsRealism scores were similar and high in both groups. Cyclist-motorists had fewer collisions with cyclists and detected cyclists at a greater distance in all situations, irrespective of cyclist visibility. Several mechanisms underlying the cognitive conspicuity of cyclists for car drivers were considered. ConclusionThe attentional selection of a cyclist in the road environment during car driving depends on top-down processing. ApplicationWe consider the practical implications of these results for the safety of vulnerable road users and future directions of research.

3D Lidar-based static and moving obstacle detection in driving environments: An approach based on voxels and multi-region ground planes

Artificial perception, in the context of autonomous driving, is the process by which an intelligent system translates sensory data into an effective model of the environment surrounding a vehicle. In this paper, and considering data from a 3D-LIDAR mounted onboard an intelligent vehicle, a 3D perception system based on voxels and planes is proposed for ground modeling and obstacle detection in urban environments. The system, which incorporates time-dependent data, is composed of two main modules: (i) an effective ground surface estimation using a piecewise plane fitting algorithm and RANSAC-method, and (ii) a voxel-grid model for static and moving obstacles detection using discriminative analysis and ego-motion information. This perception system has direct application in safety systems for intelligent vehicles, particularly in collision avoidance and vulnerable road users detection, namely pedestrians and cyclists. Experiments, using point-cloud data from a Velodyne LIDAR and localization data from an Inertial Navigation System were conducted for both a quantitative and a qualitative assessment of the static/moving obstacle detection module and for the surface estimation approach. Reported results, from experiments using the KITTI database, demonstrate the applicability and efficiency of the proposed approach in urban scenarios.

Review of pedestrian detection techniques in automotive far‐infrared video

The use of advanced driver assistance systems is becoming increasingly common in road-going vehicles. One application of these driver assistance systems is in the automated detection of vulnerable road users, such as pedestrians, using automotive far-infrared imagery. Detection of pedestrians in infrared imagery can be quite difficult because of a number of factors such as the environment, pedestrian behaviour and also the physical limitations of currently available infrared sensors. This study presents a comprehensive review of the literature currently available in the area of pedestrian detection techniques in automotive infrared imagery. The challenges associated with automated detection of pedestrians in the automotive domain are first discussed. An overview of the general structure of pedestrian detection algorithms is then presented, followed by an in-depth analysis of existing literature in the area. Some proposals for future research in the area based on the methods described in this study are then offered.