Abstract
This article presents a novel methodology for retrieving three-dimensional geometry from structures that exist only within the historical photographic record. Two shipwrecks that were raised in the early twentieth century are examined as case studies for this methodology. This paper explores the complications of using historical photographs in photogrammetric analysis and demonstrates how the shipwreck point clouds were generated using the iWitness program. In addition, it documents how photogrammetric hull data from historical photographs were scale constrained (using surviving artifacts) and used for the first time to create ship lines for academic study. A test was also conducted to demonstrate the potential accuracy of the photogrammetry program and methodology as a whole. Although the present study uses two raised shipwrecks as case studies, this methodology could easily be applied to other structures that have been “lost.”
Similar content being viewed by others
Data Availability
Not Applicable.
Code Availability
Not Applicable.
Notes
Philadelphia, Royal Savage, and Ticonderoga are not thriving, by any means. Philadelphia is suffering from ill-suited conservation strategies, Royal Savage was never conserved and it timbers were stored improperly for many decades, and Ticonderoga is suffering from a complete lack of conservation. There are, however, ongoing efforts to preserve Philadelphia and Royal Savage in their current states (Lundeberge et al. 2017; Crise et al. 2019).
The “Crown Gate” of the Dresden Zwinger, in Germany.
Not all of the images used in that project were original copies, and they also used an open-source photogrammetric software, which may have led to the subsequent inaccuracies of their model.
The fort’s lands were owned by the state of New York after 1785. The property was sold in 1820 to the Pell family, who used the grounds as their summer retreat and later built a hotel for tourists.
For definitions of terms related to nautical archaeology and ship construction, refer to “Illustrated Glossary of Ship and Boat Terms” (Catsambis et al. 2013), https://www.oxfordhandbooks.com/view/10.1093/oxfordhb/9780199336005.001.0001/oxfordhb-9780199336005-e-48.
This list of artifacts appeared on the back of each 1903 photograph of the raised sloop; the photographs are part of the Queensbury (NY) Town History Collection.
As of now, we know that a few wooden gavels, two candlesticks, and a clock were made from timbers from this vessel.
None of these measurements is reported in any of their publications concerning the wreck.
This particular photograph is not included here because the copy I have is not of high enough quality for publication.
This 35-mm focal length format is an adequate number to start with for an unknown focal length.
I used a combination of iWitnessPRO-V4 and iWitness-V4 (standard) for this project.
Zarzynski and Shaw used the insight3d program to create their model of the Tuttle Sloop.
Every image needs to have camera data assigned to it. Users may need to repeat camera assignment for each individual image as it is imported.
Users may need to zoom and pan within the images to facilitate referencing. It is recommended to use a mouse with a scroll wheel to aid in these navigational moves. The scroll wheel can be used to zoom in and out on a photograph, and when clicked and held down, it can also be used to pan the image (when dragging the mouse). Alternatively, users can manipulate the slider bars to pan within each image and can hold down “ALT” while dragging a marquee box to zoom to a designated area of the image.
Users should make sure that the initial orientation reference points are distributed around the images (not bunched together or in a single dimensional plane). Also, when using a limited number of historical images, it is important to reference more, rather than fewer, points to ensure the most accurate geometry. Users should save their work often, especially during the referencing process.
Since RMS is the accuracy of the triangulation of the 3D points in the model, this error value is calculated based on the amount of pixel deviation.
I chose three different cameras to demonstrate the repeatability of this methodology regardless of camera type.
To reiterate, this accuracy test is capable of showing only how accurate the methodology is under favorable conditions; it is not intended to falsely claim that all projects should expect these results. This accuracy test would be of little use if it were to demonstrate the obvious—that poor-quality photos generate poor results.
This was done to demonstrate that a single point cloud could be generated from photographs taken by different cameras, a scenario that is common when working with raised shipwrecks and small collections of photographs.
Although it would have been easy to check the accuracy of a photogrammetric point cloud and modify the point selection and matching to get subcentimeter accuracy for this publication, I made a conscious effort to not do that. After point selection and matching was complete and the 3D coordinates were exported, the geometry of the photogrammetric point clouds was not altered. The purpose of this accuracy test is to show an honest demonstration of the iWitness program and the capability of my methodology in a “blind” test.
Theories of vessel design and construction were used to identify the timbers that were “out of place” and would cause an implausible reconstruction. The timber locations were altered to align with plausible vessel design theory.
The final results of the point cloud scaling and a full analysis of each of these vessel’s construction and design will be covered in the author’s dissertation and subsequent publications.
I have chosen to include imperial values whenever they concern measurements related to British colonial vessels, since they were originally built using this system.
Users may choose to print the exported views and draft the hull lines by hand, or use a digital drafting software with a layers function.
These line extrusions and estimations are determined by indirect archaeological evidence on the vessel, research of other contemporary vessels, archival research, and naval architecture theory.
A common way to determine tons burthen of a vessel is to divide the product of the length of the keel, breadth, and depth below the deck by 94 (Falconer 1784).
Since the vessel’s superstructure does not exist in the photographic record, the height of the deck must be estimated (based on other vessels of this size and time period).
References
Agisoft (2020). Metashape, version 1.6.3, St. Petersburg, Russia. https://www.agisoft.com/.
Amherst Papers (1759–1763). Sections 34/42, 51, 64, and 65, Public Records Office, War Office Records, Kew, London.
Autodesk (2018). AutoCAD. https://www.autodesk.com/.
Bellico RP (2010) Empires in the mountains: French and Indian War Campaigns and Forts in the Lake Champlain, Lake George, and Hudson River Corridor. Purple Mountain Press, Fleishmanns
Bishop DE (2021) The Sloop Boscawen: hull construction and design during the mid-eighteenth century in the Champlain Valley. Doctoral Dissertation, Texas A&M University, College Station
Bitelli G, Girelli VA, Marziali M, Zanutta A (2007). Use of historical images for the documentation and the metrical study of cultural heritage by means of digital photogrammetric techniques. In XXI International CIPA Symposium, Athens, Greece
Bossom AC (ca. 1925) The restoration of Fort Ticonderoga or Fort Carillon, New York State, U.S.A. Vols I and II, located at Thompson-Pell Research Center (on loan from Amherst College Library, Archives and Special Collections)
Catsambis A, Ford B, Hamilton DL (eds) (2013) Illustrated glossary of ship and boat terms. In: The Oxford handbook of maritime archaeology. Oxford University Press, New York. https://www.oxfordhandbooks.com/view/10.1093/oxfordhb/9780199336005.001.0001/oxfordhb-9780199336005-e-48; adapted from Steffy JR (1994) Illustrated glossary of ship and boat terms. In: Wooden ship building and the interpretation of shipwrecks. Texas A&M University Press, College Station, pp. 266–98
CloudCompare (2020) 3D point cloud and mesh processing software open source project. https://www.danielgm.net/cc/.
3H Consulting Ltd. (2020). Site Recorder 4, 3H Site Recorder, Plymouth, United Kingdom. https://www.3hconsulting.com/.
Crise J, Ford B, Schwarz G (2019) From ship to kindling to ship: the digital reconstruction of the royal savage timber assemblage. In: Allen R, Ford B (eds) New life for archaeological collections. Lincoln, University of Nebraska Press
Crisman KJ (1988) Struggle for a Continent: Naval Battles of the French and Indian Wars. In: Bass GF (ed) Ships and shipwrecks of the Americas. Thames & Hudson, London, pp 129–148
Dunne KP (n.d.) The 35th regiment of foot and the British artillery at the siege of Fort William Henry and the role of Lord Loudoun, James Campbell, August 1757. https://www.academia.edu/30884306/The_35th_Regiment_of_Foot_and_the_British_Artillery_at_the_Siege_of_Fort_William_Henry_and_the_Role_of_Lord_Loudoun_James_Campbell_August_1757. Accessed 25 April 2020
Falconer W (1784) An Universal dictionary of the marine: or, a copious explanation of the technical terms and phrases employed in the construction, equipment, furniture, machinery, movements, and military operations of a ship. London
GIMP (2020) GNU image manipulation program, version 2.10.20. https://www.gimp.org/
Improvement of Beach at Lake George (1920) Warrensburg News, Warrensburg, New York
Ioannides M, Hadjiprocopi A, Doulamis N, Doulamis A, Protopapadakis E, Makantasis K, Julien M (2013) Online 4D reconstruction using multi-images available under open access. ISPRS Photogramm Remote Sens Spatial Inf Sc 2:169–174
Kane AI, Sabick CR (2002) Lake Champlain underwater cultural resources survey, Volume IV: 1999 Results and Volume V: 2000 Results. Lake Champlain Maritime Museum, Vergennes
Lewis DM (1983) The naval campaign of 1759 on Lake Champlain. Bull Fort Ticonderoga Museu 14(4):203–216
Lundeberg PK, Cohn AB, Jones JL (2017) A tale of three gunboats: Lake Champlain's Revolutionary War heritage. Lake Champlain Maritime Museum. Vergennes
Maiwald F, Vietze T, Schneider D, Henze F, Münster S, Niebling F (2017) Photogrammetric analysis of historical image repositories for virtual reconstruction in the field of digital humanities. Int Arch Photogramm Remote Sens Spatial Inf Sci 42:447–452
Martinez Batlle JR (2018) Digital photogrammetry of historical aerial photographs using open-source software. EarthArXiv. https://eartharxiv.org/bna95. Accessed 17 June 2020.
Perley S (ed) (1882). Diaries Kept by Lemuel Wood, of Boxwood, Salem Vol. 19
Photometrix (2018). iWitnessPRO user manual, Kew, Australia
Sevara C, Verhoeven G, Doneus M, Draganits E (2018) Surfaces from the visual past: recovering high-resolution terrain data from historic aerial imagery for multitemporal landscape analysis. J Archaeol Method Theory 25(2):611–642
Snavely N, Seitz SM, Szeliski R (2007) Modeling the world from internet photo collections. Int J Comput Vision 80(2):189–210
Zarzynski JW (2019) Ghost fleet awakened: Lake George’s Sunken Bateaux of 1758. SUNY Press, Albany
Zarzynski JW, Shaw B (2017) The tale of the dismembered 1757 Fort William Henry Shipwreck. The French & Indian War Society Newsletter. Lake George, New York, pp 1–7
Acknowledgements
I greatly appreciate Lake Champlain Maritime Museum’s Archaeological Director and close colleague, Chris Sabick, for sharing notes from previous excavations and for providing insight into the locations of original copies of shipwreck photographs. Similarly, I thank Fort Ticonderoga Museum’s Director of Collections, Miranda Peters, and Site Registrar and Archaeologist, Margaret Staudter, for allowing access to the Duke of Cumberland site and for sharing photographs and other files from within their collections. I warmly acknowledge the Queensbury (NY) Town Historian, Joan Aldous, and the Adirondack History Museum Archivist, Andrea Anesi, for allowing me to scan original copies of some of the photographs of the shipwrecks used in this study. I am indebted to my colleagues in the Anthropology Department at Texas A&M University. Particularly, I thank Drs. Kevin Crisman and Filipe Castro for their enthusiasm and continued support for this new research as well as PhD candidate Taryn Johnson for helping me take measurements for the accuracy test site. Finally, none of this research would be possible without the support of my family—I am eternally grateful for their advice and encouragement along the way.
Funding
Not Applicable.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflicts of interest
The author declares that he has no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Bishop, D.E. Reconstructing “Lost” Vessels: A Methodology for Applying Photogrammetric Techniques to Historical Photographs. J Mari Arch 16, 493–522 (2021). https://doi.org/10.1007/s11457-021-09311-2
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11457-021-09311-2