A driving visibility index for evaluating urban road driving safety based on point cloud data

Area of Science:

• Transportation Engineering
• Geospatial Analysis
• Computer Vision

Background:

• Road traffic safety is crucial for preventing accidents and supporting socioeconomic development.
• Current safety assessments rely on historical data and road design, but lack direct visibility metrics.
• Driving visibility is a key factor influencing driver perception and road safety.

Purpose of the Study:

• To develop a novel method for quantifying 3D driving visibility using LiDAR point cloud data.
• To introduce a Driving Visibility Index (DVI) for objective, visibility-based road safety evaluation.
• To establish a computational framework for assessing the driver's 3D visible space.

Main Methods:

• Acquisition and preprocessing of road point cloud data from LiDAR.
• Computation of the 3D driving visibility field from the driver's perspective.
• Development and application of the Driving Visibility Index (DVI) for safety assessment.

Main Results:

• The proposed method successfully quantifies 3D visible space and generates a driving safety map.
• The Driving Visibility Index (DVI) for Yixian Avenue ranged from 0.2 to 0.6, indicating suboptimal safety.
• Comparative analysis confirmed the robustness and accuracy of the proposed method against field data.

Conclusions:

• The novel method provides an objective and intuitive quantification of 3D visible space for traffic management.
• The Driving Visibility Index (DVI) offers a new basis for road design, traffic facility layout, and intelligent transportation systems.
• This approach enhances road traffic safety assessment by incorporating critical driving visibility factors.