In order to conduct this experiment and produce recommendations for future models, we utilised photographs of single and multiple long, thin mounted bones, as these are the most conducive to multiple object photography without overly obscuring each other. Our hypothesis followed that if certain factors, such as the number of photos, were kept consistent between models then there should be no significant difference between the processing times of multiple and single objects, however, the increased geometric complexity of the multiple objects may impact the quality of the finished models.
This project also necessitated the development of several coding ‘shortcuts’ that would streamline the experiment’s methodology and such scripts have become a staple of many of the team’s projects that rely upon large numbers of photographs and models, and even smaller projects as simple quality of life improvements.
Refining the efficiency and establishing standard operating procedures for photogrammetry within the team and externally, is of increasing importance, given its frequent application in the measurement of 3D shape from biological structures for further shape analyses.
Our results, which are currently in preparation for publication, found that with objects with thin profiles, the photography and reconstruction of multiple objects simultaneously significantly increases the efficiency of these stages, while slightly increasing the user involvement at later stages in order to separate the objects and with relatively little impact on the quality of the models themselves.
For a more comprehensive discussion of our methods and results, contact us. The related publication (in progress) will be shared upon publication.