ABSTRACT
ABSTRACT: In this paper, a study of the feasibility of airborne multisensor remote sensing systems to support traffic flow parameter estimation is presented. The primary objectives are (1) to define a concept of acquiring flow data, (2) to analyze the accuracy characteristics of the georeferencing component, (3) to calibrate a multisensor system, and (4) to present an example of the use of directly georeferenced imagery to estimate traffic flow parameters. The sensor suite that is presented is based on dual-frequency geodetic grade Global Positioning System (GPS) receivers, a high accuracy inertial measurement unit (IMU), and a 4000 by 4000 charge-coupled device (CCD) digital camera.The traffic flowparameters that are investigated are (1) turning movements and volumes based on a single intersection image sequence and (2) velocity estimation based on vehicle tracking in the image sequence. Experimental results from a helicopter test flight over a reference area, using the Ohio State University (OSU) GPS/IMU/CCD prototype system are presented.
