Automated, Targetless Photogrammetric Measurement: Potential and Limitations

Abstract

Close-range photogrammetry is now a widely-adopted tool for a host of 3D measurement tasks across areas as diverse as industrial metrology, forensics and heritage recording. Whereas with industrial and engineering applications the utilisation of targets is near universal, the adoption of newly developed automated, targetless orientation and object reconstruction is becoming increasingly more widespread in other application domains. This naturally poses the question of why so-called Structurefrom-Motion (SfM) or Feature-Based Matching (FBM) techniques for automatic photogrammetric measurement are yet to employed to any significant extent in large-scale metrology, in spite of the obviously appealing prospect of dispensing with artificial signalisation, be it through physical targets or projection schemes. This presentation reviews the development of targetless, automated photogrammetric measurement, which has been given extra impetus through the introduction of drones/UAVs as camera platforms. Prospects for adoption of SfM approaches in industrial and engineering measurement are then discussed, with particular emphasis upon the broader requirements of 3D feature point measurement. Although the potential of automated targetless photogrammetry is significant, there are nevertheless many limitations to adoption, not least in the area of accuracy, feature point identification and operational constraints applying to drones.