An official website of the United States government
Road conditions, often degraded by insufficient maintenance or adverse weather, significantly contribute to accidents, exacerbated by the limited human reaction time to sudden hazards like potholes. Early detection of distant potholes is crucial for timely corrective actions, such as reducing speed or avoiding obstacles, to mitigate vehicle damage and accidents. This paper introduces a novel approach that utilizes perspective transformation to enhance pothole detection at different distances, focusing particularly on distant potholes. Perspective transformation improves the visibility and clarity of potholes by virtually bringing them closer and enlarging their features, which is particularly beneficial given the fixed-size input requirement of object detection networks, typically significantly smaller than the raw image resolutions captured by cameras. Our method automatically identifies the region of interest (ROI)—the road area—and calculates the corner points to generate a perspective transformation matrix. This matrix is applied to all images and corresponding bounding box labels, enhancing the representation of potholes in the dataset. This approach significantly boosts detection performance when used with YOLOv5-small, achieving a 43% improvement in the average precision (AP) metric at intersection-over-union thresholds of 0.5 to 0.95 for single class evaluation, and notable improvements of 34%, 63%, and 194% for near, medium, and far potholes, respectively, after categorizing them based on their distance. To the best of our knowledge, this work is the first to employ perspective transformation specifically for enhancing the detection of distant potholes.
more »
« less
A Reliability-Aware Vehicular Crowdsensing System for Pothole Profiling
Accurately profiling potholes on road surfaces not only helps eliminate safety related concerns and improve commuting efficiency for drivers, but also reduces unnecessary maintenance cost for transportation agencies. In this paper, we propose a smartphone-based system that is capable of precisely estimating the length and depth of potholes, and introduce a holistic design on pothole data collection, profile aggregation and pothole warning and reporting. The proposed system relies on the built-in inertial sensors of vehicle-carried smartphones to estimate pothole profiles, and warn the driver about incoming potholes. Because of the difference in driving behaviors and vehicle suspension systems, a major challenge in building such system is how to aggregate conflicting sensory reports from multiple participating vehicles. To tackle this challenge, we propose a novel reliability-aware data aggregation algorithm called Reliability Adaptive Truth Discovery (RATD). It infers the reliability for each data source and aggregates pothole profiles in an unsupervised fashion. Our field test shows that the proposed system can effectively estimate pothole profiles, and the RATD algorithm significantly improves the profiling accuracy compared with popular data aggregation methods.
more »
« less
- PAR ID:
- 10294399
- Date Published:
- Journal Name:
- Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
- Volume:
- 3
- Issue:
- 4
- ISSN:
- 2474-9567
- Page Range / eLocation ID:
- 1 to 26
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
This paper focuses on the detection of cyber-attack on a communication channel and simultaneous radar health monitoring for a connected vehicle. A semi-autonomous adaptive cruise control (SA-ACC) vehicle is considered which has wireless communication with its immediately preceding vehicle to operate at small time-gap distances without creating string instability. However, the reliability of the wireless connectivity is critical for ensuring safe vehicle operation. The presence of two unknown inputs related to both sensor failure and cyber-attack seemingly poses a difficult estimation challenge. The dynamic system is first represented in descriptor system form. An observer with estimation error dynamics decoupled from the cyber-attack signal is developed. The performance of the observer is extensively evaluated in simulations. The estimation system is able to detect either a fault in the velocity measurement radar channel or a cyber-attack. Also, the proposed observer-based controller achieves resilient SA-ACC system under the cyber-attacks. The fundamental estimation algorithm developed herein can be extended in the future to enable cyber-attack detection in more complex connected vehicle architectures.more » « less
-
Artificial intelligence (AI) supported network traffic classification (NTC) has been developed lately for network measurement and quality-of-service (QoS) purposes. More recently, federated learning (FL) approach has been promoted for distributed NTC development due to its nature of unshared dataset for better privacy and confidentiality in raw networking data collection and sharing. However, network measurement still require invasive probes and constant traffic monitoring. In this paper, we propose a non-invasive network traffic estimation and user profiling mechanism by leveraging label inference of FL-based NTC. In specific, the proposed scheme only monitors weight differences in FL model updates from a targeting user and recovers its network application (APP) labels as well as a rough estimate on the traffic pattern. Assuming a slotted FL update mechanism, the proposed scheme further maps inferred labels from multiple slots to different profiling classes that depend on, e.g., QoS and APP categorization. Without loss of generality, user profiles are determined based on normalized productivity, entertainment, and casual usage scores derived from an existing commercial router and its backend server. A slot extension mechanism is further developed for more accurate profiling beyond raw traffic measurement. Evaluations conducted on seven popular APPs across three user profiles demonstrate that our approach can achieve accurate networking user profiling without invasive physical probes nor constant traffic monitoring.more » « less
-
The Wire Flyer towed vehicle is a new platform able to collect high-resolution water column sections. The vehicle is motivated by a desire to effectively capture spatial structures at the submesoscale. The vehicle fills a niche that is not achieved by other existing towed and repeat profiling systems. The Wire Flyer profiles up and down along a ship-towed cable autonomously using controllable wings for propulsion. At ship speeds between 2 and 5 kt (1.02–2.55ms21), the vehicle is able to profile over prescribed depth bands down to 1000 m. The vehicle carries sensors for conductivity, temperature, depth, oxygen, turbidity, chlorophyll, pH, and oxidation reduction potential. During normal operations the vehicle is typically commanded to cover vertical regions between 300 and 400min height with profiles that repeat at kilometer spacing. The vertical profiling speed can be user specified up to 150mmin21. The high-density sampling capability at depths below the upper few hundred meters makes the vehicle distinct from other systems. During operations an acoustic modem is used to communicate with the vehicle to provide status information, data samples, and the ability to modify the sampling pattern. This paper provides an overview of the vehicle system, describes its operation, and presents results from several cruises.more » « less
-
The Wire Flyer towed vehicle is a new platform able to collect high-resolution water column sections. The vehicle is motivated by a desire to effectively capture spatial structures at the submesoscale. The vehicle fills a niche that is not achieved by other existing towed and repeat profiling systems. The Wire Flyer profiles up and down along a ship-towed cable autonomously using controllable wings for propulsion. At ship speeds between 2 and 5 kt (1.02–2.55 m s−1), the vehicle is able to profile over prescribed depth bands down to 1000 m. The vehicle carries sensors for conductivity, temperature, depth, oxygen, turbidity, chlorophyll, pH, and oxidation reduction potential. During normal operations the vehicle is typically commanded to cover vertical regions between 300 and 400 m in height with profiles that repeat at kilometer spacing. The vertical profiling speed can be user specified up to 150 m min−1. The high-density sampling capability at depths below the upper few hundred meters makes the vehicle distinct from other systems. During operations an acoustic modem is used to communicate with the vehicle to provide status information, data samples, and the ability to modify the sampling pattern. This paper provides an overview of the vehicle system, describes its operation, and presents results from several cruises.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1145/3369815
