Roadside LiDAR (Light Detection and Ranging) sensors are recently being explored for Intelligent Transportation System aiming at safer and faster traffic management and vehicular operations. However, massive data volume, occlusion, and limited viewing angles are significant obstacles to the widespread use of roadside LiDARs. In this dissertation, we address three major challenges to enable applications of Intelligent Transportation System through roadside LiDAR data: (i) real-time transmission of the massive point-cloud data from the roadside LiDAR devices to the cloud using 5G network, (ii) mitigating sensor occlusion problem to increase coverage and detect events occurred in occluded regions of a sensor, and (ii) summarizing the roadside LiDAR dataset.
We develop a novel grouped wavelet technique for Static roadside LiDAR data compression (i.e., SLiC) so that the data can be streamed to the cloud in real-time. Experimental results show that SLiC can compress up to 1.9$\times$ more effectively than the state-of-the-art compression method can do. Moreover, SLiC is computationally more efficient in achieving 2$\times$ improvement in bandwidth usage over the best alternative. Even with this impressive gain in communication and storage efficiency, SLiC retains down-the-pipeline application's accuracy. We also address the sensor occlusion problem by proposing a calibration technique to fuse multiple LiDARs. We show that our technique works on various data granularity and enables real-time analytics for roadside traffic monitoring. We conduct evaluations using a wide range of simulated traffic scenarios, demonstrating that fusion reduces occlusion and enhances the accuracy of vehicle counting and near-collision detection. We apply our algorithm to real traffic data and demonstrate utility in classifying vehicles and detecting occluded traffic participants. Finally, we summarize the roadside LiDAR data by transforming the point cloud data into time series and proposing a new pattern mining task that considers joining a set of time series based on an arbitrary number of subsequences (i.e., patterns) with arbitrary lengths. We develop a fast algorithm named MultiPAL to join multiple time series at interactive speed to identify valuable events from the roadside LiDAR data.
Roadside LiDAR, Compression, LiDAR Calibration, Data Fusion, Time series, Pattern Mining
Level of Degree
Department of Computer Science
First Committee Member (Chair)
Second Committee Member
Third Committee Member
Fourth Committee Member
National Science Foundation (NSF), NEC Laboratories America, Inc.
Mollah, Md Parvez. "Roadside LiDAR Data Processing for Intelligent Transportation System." (2023). https://digitalrepository.unm.edu/cs_etds/121