Capturing the Past for the Future: an Evaluation of the Effect of Geometric Scan Deformities on the Performance of Aerial Archival Media in Image‐based Modelling Environments

Abstract

Image-based modelling (IBM) applications can be used to successfully extract two-and three-dimensional information from historic aerial and satellite imagery for use in archaeological and other landscape-oriented research. The robust and highly automated nature of these applications can allow for image matching and dense scene reconstruction, even when input images lack information necessary for processing in traditional photogrammetric workflows. However, the overall quality of the content derived from IBM applications can be directly influenced by a number of variables, including the manner in which archival content is digitized. Significant geometric distortion can be induced in the scanning process when using non-photogrammetric scanning devices, which can have a negative effect on the subsequent processing of scanned imagery and accuracy of derived content such as orthomosaics and digital elevation models (DEMs). While this is a well-known issue for users of high end photogrammetry packages, the advent of IBM applications has brought the use of photogrammetric-style techniques to a wider audience who may not be aware of the impact that geometric errors of this nature can have on the quality of derived data products. This issue is demonstrated here using a set of degraded historic World War II (WWII) Allied reconnaissance photographs that have been digitized on both photogrammetric and nonphotogrammetric platforms and subsequently processed in a well-known IBM application using identical processing parameters. The results are compared to those from a recently collected set of vertical stereopairs and an airborne laser scan derived DEM.

Joining multi-epoch archival aerial images in a single SfM block allows 3-D change detection with almost exclusively image information

“…This archive was recently released and is freely available on the IGN's website (https://remonterletemps.ign.fr/). The fact that these images were scanned with photogrammetric scanners is important for further photogrammetric processing as noted by Sevara (2016) and observed in a preliminary study (Feurer et al, 2017). We downloaded a sample of these data at four different epochs allowing for a stereo coverage of the entire test site (coloured polygons on Figure 1).…”

“…In their pioneering work, Chandler and Cooper (1989) reminded that the interior orientation consists of two groups of parameters: the parameters that characterise the displacement of the principal point relative to the fiducial marks and the parameters that characterise the lens distortion. To these parameters, it is necessary to add the deformation due to the scanning, which has a strong impact on the performance of the 3-D estimation (Sevara, 2016). These additional unknowns severely increase the need for ground control points, which in turn raises new issues : the need for expensive additional ground survey and the problem of the accuracy of the orientation estimation with self-calibration algorithms (Aguilar et al, 2013).…”

Engaging with the Canopy—Multi-Dimensional Vegetation Mark Visualisation Using Archived Aerial Images

“…First, every photograph suffers from distortions induced by the camera optics. These are, however, complemented by possible film deformations-resulting from film shrinkage and/or expansion due to their storage in a non-optimised environment-as well as serious geometric errors that can be introduced at the scanning stage [61]. Moreover, the focal length of the images is unknown, which means that the SfM step does not have a suitable initial value to start from.…”

Analysis of Changes in Forest Structure using Point Clouds from Historical Aerial Photographs

“…Distortion parameters are another unknown aspect of this problem. In the case of these parameters, it is necessary to take into consideration errors resulting from the scanning of negatives [38]. These additional factors significantly increase the necessity to use a large number of ground control points.…”