Mobiusi/Tunnel-and-Special-Scenarios-Traffic-Safety-Dataset

Tunnel and Special Scenarios Traffic Safety Dataset

The current transportation industry is facing increasingly severe safety hazards, especially in tunnels and special scenarios where traffic accidents frequently occur and there is an urgent need for effective monitoring and evaluation methods. Existing monitoring systems often rely on manual monitoring, which is inefficient and prone to errors, lacking real-time and accuracy. This dataset aims to support the training of object detection algorithms by providing high-quality image data to achieve intelligent traffic safety monitoring. Data collection is performed using high-resolution cameras in different tunnels and special scenarios to ensure coverage of various situations. Additionally, a multi-round annotation and expert review mechanism is adopted to ensure consistency and accuracy in annotations. The data is stored in JPG format and organized by scenario for convenient use and analysis. The core advantage of the dataset is its high-quality annotation, with all image annotations achieving over 95% accuracy and a consistency check rate up to 98%. The new data enhancement techniques used have improved the model's generalization capability by 20%, effectively reducing the incidence of traffic accidents and enhancing safety in practical applications.

Technical Specifications

Field	Type	Description
file_name	string	File name
quality	string	Resolution
vehicle_count	int	The total number of vehicles detected in the image.
traffic_density	float	The density of vehicles in the image.
object_type	string	The type of object detected in the image, such as vehicles, pedestrians, etc.
lane_occupation	string	The status of lane occupation as observed in the image.
visibility	string	The visibility conditions in the image, which may include sunny, foggy, rainy, etc.
scene_type	string	The type of scene captured in the image, such as tunnel, mountainous area, urban, etc.
road_condition	string	The condition of the road as shown in the image, such as dry, slippery, snowy, etc.
lighting_condition	string	The lighting condition at the time the image was taken, such as daytime, nighttime, dusk, etc.
incident_presence	boolean	Whether a traffic incident is present in the image.
speed_limit_signs	string	The speed limit signs detected in the image.

Compliance Statement

Authorization Type	CC-BY-NC-SA 4.0 (Attribution–NonCommercial–ShareAlike)
Commercial Use	Requires exclusive subscription or authorization contract (monthly or per-invocation charging)
Privacy and Anonymization	No PII, no real company names, simulated scenarios follow industry standards
Compliance System	Compliant with China's Data Security Law / EU GDPR / supports enterprise data access logs

Source & Contact

If you need more dataset details, please visit Mobiusi. or contact us via contact@mobiusi.com