Skip to main content

An assessment of archive stereo-aerial photographs for 3-dimensional reconstruction of damaged and destroyed archaeological earthworks.

Papworth, H. E., 2014. An assessment of archive stereo-aerial photographs for 3-dimensional reconstruction of damaged and destroyed archaeological earthworks. Doctoral Thesis (Doctoral). Bournemouth University.

Full text available as:

Papworth,Heather Elizabeth_Ph.D._2014.pdf



Archaeological earthworks are being damaged and destroyed at a rate and scale never before seen, which has resulted from increased mechanisation of human activity in the landscape since World War II. Along with natural degradation processes, recording earthwork metrics prior to their loss is increasingly difficult, which can subsequently hinder the interpretation of a site or landscape because of this missing evidence. A tool for regaining such data is vital to alleviate this problem and to fulfil the stipulation for metric information as required by national and international conservation charters. This research investigates whether it is possible to regain earthwork metrics from archive stereo-aerial photographs (SAPs) using digital photogrammetry to create digital surface models (DSMs) of archaeological sites within the UK dating from the 1940s to 2010. A literature search confirmed the utility of SAPs for reconstructing geomorphological events, such as landslides, whilst also verifying that such an approach had not been thoroughly investigated for archaeological adaptation. Via experimentation, a photogrammetric workflow has been designed and a number of variables identified that affect the quality of DSMs obtained from SAPs. The magnitude of these variables has been verified by quantitative assessment using independent survey data, namely Airborne Laser Scanning (ALS) gathered by the Environment Agency, and ground-based collection using Global Navigation Satellite Systems (GNSS) and Terrestrial Laser Scanning (TLS). Empirical differences between these independent data and the SAP DSMs were identified using global statistical measures such as Mean Error (ME), Standard Deviation (SD) and root mean square error (RMSE), and spatial autocorrelation techniques, namely Local Moran’s I. Two study sites were selected on which to ascertain whether variations occur in the empirical quality of SAP DSMs and archaeological content at different locations. Over six decades of photography were collected for Flowers Barrow Hillfort, situated near Lulworth in Dorset, UK, which has remained in good condition throughout this period, due to the protection afforded it by inclusion within Ministry of Defence land. Eggardon Hillfort and earthworks, near Bridport in Dorset, UK, were also selected due to the exceptional preservation state of some earthworks, versus the plough-damaged remains of others. These sites thus offered an opportunity to rigorously test the reconstruction capabilities of the SAPs. The results from both study sites confirmed that the metric quality of SAP DSMs improves as the age of the imagery decreases, although this is dependent on image quality, scanner properties (i.e. whether the scanner is photogrammetric or desktop) and the result of the block bundle adjustment in the photogrammetric software. This thesis concludes that SAPs can recreate earthwork metrics and provides a list of considerations for archaeologists to consult when planning the use of SAPs for creating DSMs. Recommendations for future work are provided that encourage the investigation of SAPs from other countries and the rigorous assessment of DSMs derived from structure-from-motion (SfM) software that is rapidly gaining popularity.

Item Type:Thesis (Doctoral)
Additional Information:If you feel that this work infringes your copyright please contact the BURO Manager.
Uncontrolled Keywords:Photogrammetry, Archive Stereo-Ariel Photography, Laser Scanning, Archaeology, Digital Surface Model.
Group:Faculty of Science & Technology
ID Code:21780
Deposited By: Symplectic RT2
Deposited On:26 Mar 2015 11:56
Last Modified:09 Aug 2022 16:03


Downloads per month over past year

More statistics for this item...
Repository Staff Only -