Premium
An automated approach to the analysis of the arrangement of post‐pits at Stanton Drew
Author(s) -
Linford P. K.
Publication year - 2005
Publication title -
archaeological prospection
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.785
H-Index - 38
eISSN - 1099-0763
pISSN - 1075-2196
DOI - 10.1002/arp.252
Subject(s) - concentric , hough transform , artificial intelligence , interpretation (philosophy) , computer science , set (abstract data type) , algorithm , mathematics , geometry , geology , image (mathematics) , programming language
After the surprising discovery of concentric arrangements of post‐pits within the megalithic stone circles at Stanton Drew in Somerset, it was decided to map their geometry using an automated approach. The algorithm developed relies on the use of correlation registration to allow the locations of pit anomalies to be identified automatically. Concentric circular arrangements of pits are then classified using a modified Hough transform, adapted to detect circular rather than linear patterns. Finally the co‐ordinates of the centres of all pits identified as lying on the same circle are extracted and fed to an algorithm that determines the parameters of the circle of best fit by the method of least squares. The result is an objective determination of the parameters of the nine concentric circles of pits within the Great stone circle and three circles within the south‐southwest stone circle, which it is hoped will help inform the interpretation of the monument. Although the computational procedures described represent a solution to the analysis of a possibly unique data set, it is hoped that elements of the method might provide ideas for techniques that can be generalized for other applications. Copyright © 2005 John Wiley & Sons, Ltd.