Friday 26 August 2005

G6
1000-1200 hours

497
Vehicle classification and traffic flow estimation from airborne LiDAR/CCD data
Grejner-Brzezinska, Dorota¹, Toth, Charles¹, Moafipoor, Shahram¹, Paska, Eva¹
1 The Ohio State University, Satellite Positioning and Inertial Navigation (SPIN) Lab, Columbus, OH, USA
Author email: dbrzezinska@osu.edu
This paper provides a summary review of a 3-year research program on the feasibility of using airborne LiDAR and imagery collected simultaneously over transportation corridors for estimation of traffic flow parameters, such as (1) vehicle counts, (2) vehicle classification, (3) velocity estimates based on vehicle categories and using sensor navigation data, and (4) intersection movement patterns. This work is conducted by The National Consortium for Remote Sensing in Transportation-Flows, led by OSU, supported by the U.S. Department of Transportation and NASA. The major focus is on improving the efficiency of transportation systems by integration of remotely sensed data with traditional ground data to monitor and manage traffic flows. Recent enhancements in spatial and temporal resolution of LiDAR data can now allow for effective detecting and tracking of moving objects. Thus, in this paper, special emphasis is on LiDAR-based approach; methodology of extracting vehicle information together with the road surface modeling with precisely georeferenced LiDAR data, augmented by LiDAR intensity information are discussed. We demonstrate that intelligent algorithms that we developed are capable of fast and robust identification of the vehicle shapes (especially the vertical profiles), proving LiDAR's ability to preserve the geometry of a moving object better, as compared to conventional image projection, where it can be significantly distorted. We demonstrate that if LiDAR data of sufficient spatial density are available, vehicle extraction and their classification can be effectively performed in parallel to the efficient and automated road surface extraction and modeling. It is shown, however, that for better accuracy and reliability, a fusion of LiDAR with frame image data is desirable. The actual example of the vehicle extraction and the road surface modeling will be based on the high-density (2-4 points/m2) LiDAR data collected on February 19, 2004 over downtown Toronto area with the Optech ALTM 30/70 LiDAR system.

Return to Oral Presentations