AGOCG logo
Graphics Multimedia VR Visualisation Contents
Training Reports Workshops Briefings Index
Also available as an Acrobat File Back Next

Editorial

Abstract

Introduction

Methodology

Results
   1:10,000
   results1.html#4.2
   Image Analysis

Conclusions

Acknowledgements

References


Case Studies Index

The Application of Digital Photogrammetric Techniques and Aerial Photography to the Preservation of Archaeological Detail.

4.3 Image Analysis and spatial data measurement

As a final stage in this study, thought was given to researchers who may have temporary access to the DPW but generally carry out research on desktop systems. These systems, whilst powerful, lack the processing capabilities of the DPW so routine 3D spatial analysis and dataset comparison can be slow. To illustrate the use of such systems in manipulating DPW-based data, the orthophotos generated from the 1:2000 scale images were exported to a standard image analysis package (HL Image++). It is possible to carry out image analysis of DPW orthophotos on any desktop system that runs imaging software capable of calibration. An area of the DPW orthophoto of Circle B was exported to a PC system and the 2D images of the stones examined. Blob analysis of the image rapidly identified and outlined the individual stones and the spatial measurements of each were determined by the system. In all, this procedure takes less than 5 minutes on a mid-range Pentium PC with 32MB RAM. The output from the analysis is a large dataset including stone axis lengths, perimeter lengths and positional measurements that lend themselves to graphical presentation. Figure 8A illustrates the images exported and processed using HLImage++. Figure 8B illustrates some useful parameters of the stones generated by image analysis that can be visually represented on standard plots.


Figure 8. 2D Image analysis of a section of Circle B.
A. Transferred image after blob analysis and identification of individual stones. The plot shows the identified centroids of each stone and is recorded for future comparison. This allows identification of any stones moved by human intervention, cattle, storm processes etc.
B. Sample data plots from spatial measurements generated within a standard image analysis package (in this case, HLImage ++). The total area histogram sh ows a sample set that has a fairly even spread of stone dimensions. As decay proceeds, the mean will shift towards smaller values and the histogram should become progressively more skewed or polymodal. The roundness/ perimeter plot facilitates a simple representation of the degree of roundness of the stones, a characteristic produced by certain forms of weathering.

Graphics     Multimedia      Virtual Environments      Visualisation      Contents