Extracting dynamic spatial data from airborne imaging sensors to support traffic flow estimation

Abstract

The recent transition from analog to totally digital data acquisition and processing techniques in airborne surveying represents a major milestone in the evolution of spatial information science and practice. On one hand, the improved quality of the primary sensor data can provide the foundation for better automation of the information extraction processes. This phenomenon is also strongly supported by continuously expanding computer technology, which offers almost unlimited processing power. On the other hand, the variety of the data, including rich information content and better temporal characteristics, acquired by the new digital sensors and coupled with rapidly advancing processing techniques, is broadening the applications of airborne surveying. One of these new application areas is traffic flow extraction aimed at supporting better traffic monitoring and management. Transportation mapping has always represented a significant segment of civilian mapping and is mainly concerned with road corridor mapping for design and engineering purposes, infrastructure mapping and facility management, and more recently, environmental mapping. In all these cases, the objective of the mapping is to extract the static features of the object space, such as man-made and natural objects, typically along the road network. In contrast, the traffic moving in the transportation network represents a very dynamic environment, which complicates the spatial data extraction processes as the signals of moving vehicles should be identified and removed. Rather than removing and discarding the signals, however, they can be turned into traffic flow information. This paper reviews initial research efforts to extract traffic flow information from laserscanner and digital camera sensors installed in airborne platforms.