Identification of Weather Conditions Related to Roadside LiDAR Data

Traffic data collection is essential for traffic safety and operations studies and has been recognized as a fundamental component in the development of intelligent transportation systems. In recent years, growing interest is shown by both industrial and academic communities in high-resolution data that can portray traffic operations for all transportation participants such as connected or conventional vehicles, transit buses, and pedestrians. Roadside Light Detection and Ranging (LiDAR) sensors can be deployed to collect such high-resolution traffic data sets. However, LiDAR sensing could be negatively affected in the context of rain, snow, and wind conditions as the collected 3D point clouds of surrounding objects may drift. Weather-caused impacts can lead to difficulties in data processing and even accuracy compromise. Consequently, solutions are desired and sought, focused on the issue that the affected data have been identified through a labor-intensive and time-consuming process. In this research, a methodology is proposed for developing an automatic identification of the LiDAR data sets that are affected by rain, snow, and wind conditions. First, the impacts of rain, snow, and wind are characterized using statistical measures. Detection distance offset (DDO) and Detection distance offset for wind (DDOW) are calculated and investigated, and it shows that rain or snow conditions can be differentiated according to the standard deviation of the DDOs. Snow conditions can be additionally identified using the sum of the DDOs. Unlike rain and snow, wind conditions can be recognized by the differences between the upper and lower boundaries of DDOs, and therefore, a separate analysis is developed. Based upon the multiple analyses developed in the research, an automatic identification process is designed. The thresholds for identifying rain, snow, and wind conditions are set up, respectively. The process is validated using realistic roadside LiDAR data collected at the intersection of McCarran Blvd and Evans Ave in Reno, Nevada. The validation demonstrated that the proposed identification could precisely detect affected data sets in the context of rain, snow, and wind conditions.

Level thesis:degree_level

Master's Degree

Year dc:date.issued

2021

Author dc:creator

• Tian, Yuxin

Advisor dc:contributor.advisor

• Xu, Hao

Committee members dc:contributor.committeemember

• Tian, Zong
• Kelley, Richard

• Automatic identification Method
• Detection Distance Offset
• Detection Distance Offset for Wind
• DIFFSUM
• Inclement Weather Conditions
• Roadside LiDAR

Statement dc:rights

• Creative Commons Attribution-NonCommercial-ShareAlike 4.0 United States

Handle dc:identifier.uri

http://hdl.handle.net/11714/7817

OAI identifier oai:identifier

oai:scholarwolf.unr.edu:11714/7817