779 
CHALLENGES OF PHOTOGRAMMETRIC INTRA-ORAL TOOTH MEASUREMENT 
H. L. Mitchell *• •, R.G. Chadwick b 
3 Civil, Surveying & Environmental Engineering, University of Newcastle, Australia - harvey.mitchell@newcastle.edu.au 
b Dental School, University of Dundee, United Kingdom - r.g.chadwick@dundee.ac.uk 
KEY WORDS: Medicine, Dentistry, Automation, Application, Close range, Teeth 
ABSTRACT: 
Recording the surface shape of a living tooth would seem to be a straightforward photogrammetric task, but the researchers’ 
experiences in developing a procedure for routine intra-oral tooth measurement shows that, in practice, photogrammetric 
measurement of the living tooth in the mouth has difficult challenges which are predominantly photographic. Although the problem 
of access can be overcome by using specialist intra-oral dental cameras, arranging multiple images is problematic. A second 
challenge arises because of the optical characteristics of dental enamel: the tooth surface is featureless and is unsuitable for 
photogrammetric mapping without some augmentation, and the enamel is also highly reflective. Some imaging tests have been 
carried out on teeth in the mouth in the search for an automated measurement technique, but, to avoid patient discomfort, further 
investigations have been carried out on an extracted tooth, whose characteristics are similar to those of the live tooth. In an example 
cited here, a pair of images of an extracted frontal incisor was collected with a single camera, a camera base of about 6 mm and a 
base-to-height ratio of 1:2. To make the enamel surface both opaque and textured, it was painted with a weak water colour solution. 
Features detected with an interest operator were matched by area-based matching, and coverage was of acceptable accuracy and 
density except in areas of illumination glare. The measurement should be repeatable with living teeth by duplicating the conditions 
of the photography, but procedures with living patients are noticeably more awkward than working with inert objects. 
1. INTRODUCTION 
1.1 General 
The ultimate goal of the work outlined here is to measure living 
teeth within the human mouth using automated 
photogrammetric methods. The work has begun with attempts 
to measure extracted teeth, as this clearly avoids the need for 
uncomfortable access to patients during continual 
experimentation. Recording the surface shape of a tooth would 
seem to be a straightforward photogrammetric task, but the 
researchers have found that, in practice, photogrammetric 
measurement of the living tooth in the mouth, and even the 
extracted tooth, has a number of distinct and difficult 
photographic challenges. This paper reports some of the 
experiences and progress of the researchers in initial attempts at 
developing a procedure for routine automated intra-oral tooth 
measurement. 
1.2 Dental studies and uses of surface measurement 
Measurement of the shape of living human teeth has beneficial 
applications in dental research, to enable researchers to quantify 
the loss of tooth surface material. This loss can be brought 
about by contact with acids produced by bacteria upon 
metabolising sugars in foods (causing dental caries/decay) or by 
direct contact with the tooth surface itself (dental erosion). It 
may also arise from mechanical wear and tear such as seen in 
abrasion or attrition. Considerable dental research effort is 
expended investigating each of these matters. The 
investigations into all of these issues frequently involves 
measuring teeth shapes at various epochs, in order to carry out 
comparisons which indicate the volume and distribution of 
decayed, eroded or abraded material. The studies can be 
expected to involve the measurement of statistically significant 
numbers of teeth. Hence, it is arguable that dental research into 
tooth surface loss represents a realistic area of demand for 
measurement, and it is foreseeable that it could be undertaken 
photogrammetrically. 
In addition, measurement of the shape of living teeth has 
applications for clinicians, primarily for the fabrication of 
crowns, as is borne out by the existence of commercial optical 
(but not photogrammetric) measuring systems (e.g., Sirona 
GmbH, 2008; D4D Technologies, 2008). The use of shape 
measurement for recording patients’ dental histories is also 
imaginable, although this is not seen to be occurring at the 
moment. 
It is assumed by the writers of this paper that the convenience 
requirements of a commercial measurement system for clinical 
use would be much more stringent than for research use, so it is 
the research use, including the writers’ own dental research use, 
(e.g., Chadwick et al., 2005) to which these studies are 
currently targetted. 
1.3 Current methods 
Despite the extensive need for tooth measurement, direct intra 
oral measurement can currently be prohibitively expensive. 
Consequently, measurement of the three-dimensional shape of 
teeth (whether for research or clinical use) normally proceeds 
by taking castings of the teeth and, most often, by the formation 
of replicas from the castings. The replicas are then measured, 
often by mechanical methods, using styli, or just occasionally 
by optical/imaging methods, including photogrammetry 
(Grenness et al., 2008). To obtain statistically meaningful 
quantities of patient tooth measurements in an extensive 
investigation of dental erosion between 2001 and 2003, the 
Corresponding author.