Real-Time Visualization Method for Estimating 3D Highway Sight Distance Using LiDAR Data
Yang Ma,Yubing Zheng,Jianchuan Cheng,Said Easa
DOI: https://doi.org/10.1061/jtepbs.0000228
2019-01-01
Journal of Transportation Engineering Part A Systems
Abstract:Light detection and ranging (LiDAR) data provide a rather precise depiction of the real three-dimensional (3D) road environment and have been used by some researchers to produce more precise available sight distance (ASD) results compared with those obtained based on conventional digital elevation models with low resolution. However, existing methods have some difficulties in creating digital surface models to accurately estimate ASD using LiDAR data. In addition, dynamic visualization of the driver's visual conditions along the highway throughout ASD assessment (which is important for monitoring the results in real time) has not been achieved by existing studies. To fill these gaps, this paper discusses the development of a new procedure supported by MATLAB for evaluating, in a real-time visualization manner, ASD along an existing highway based on LiDAR data. With an innovative algorithm that combines cylindrical perspective projection and modified Delaunay triangulation, the computation is processed in real time along the vehicle trajectory, which is represented by a set of points, whereas the driver's successive perspective views and sight distance results are generated simultaneously. A comparative case study is presented to demonstrate that the new method is more accurate than conventional methods and more flexible for evaluating ASD along highways with complicated roadside components.