care
a from
s based
more
not be
to an
ernally
uitable
both a
curate
1e care
lysis.
tment-
res of
1988).
using
metry.
R.R.,
-time
metry,
ociety
0-304.
stment
1etry.
range
18 and
ort on
uction
Non-
arara,
metry
CCD or not CCD
D. R. Gordon
Civil Engineering Department
University of Canterbury
Christchurch, NEW ZEALAND
EMAIL: Gordon@civl.canterbury.ac.nz
ABSTRACT
An ongoing Biostereometric project aims to produce customised shoe-making lasts for people whose feet are
too unusual to be fitted from normal (or abnormal) shoe stocks, or even to have shoes made from standard
(or temporarily modified) lasts. This work requires some 30 to 50 points to define the total shape of such
feet, and it is particularly important that irregular and unusually shaped feet are described in sufficient detail
to ensure that the consequent last design and computer-controlled milling of individualised lasts are correct.
An overriding criteria of this programme has been to make the whole measurement and design process
available to the bespoke shoe trade, and to ensure that lasts can be produced for a reasonable price. It is
important that photogrammetrists and other (expensive) mensuration and design professionals are not
required in the normal operation of producing such lasts.
This study compares the cost and efficacy of some methods of capturing this data. In particular it considers
digital video pictures with automated and semi-automated data capture, as an alternative to conventional
photography which is hand digitised on enlarged prints.
The comparisons will concentrate on cost effectiveness, accuracy, and on ease and reliability when used by
non-technical non-photogrammetrists.
KEY WORDS: Photogrammetry, Close-Range, Biostereometrics, CCD, orthotic, shoes, feet
INTRODUCTION
I have reported elsewhere (Gordon, 1991) on my experi-
ence with a Christchurch made-to-measure footwear
company, who estimate they decline some 10 enquiries a
week from potential clients (from a total population in the
Canterbury area of some 400,000 people) because their
feet differed too much from the company's existing stock of
lasts. The Last Footwear Company regularly make tempor-
ary modifications to their lasts for individual clients, usually
by adding suitably shaped pieces of leather, but that there
are limits to what they attempt in this regard.
Dr Ken Whybrew of the Mechanical Engineering Depart-
ment in the Engineering School of the University of
Canterbury, has joined me in a project to acquire dimen-
sional data on a (misshapen) human foot and use the data
(after due processing) to control a milling machine in
producing a last from which a shoe can be made to fit that
foot. The concept is not new (Duncan, et al 1974), and we
are agreed that while neither the measuring nor the milling
present insuperable problems, the data processing to link
them will require more work.
My additional aim for this project is that the whole process
should be available to, and operated by, the cobblers who
make shoes for their clients. All of the numerical process-
ing required for the data collection through to the produc-
tion of the control data for the milling machine can be
done on a personal computer of quite modest power and
capacity. My aim is to have a turnkey process in which the
problems in data acquisition, data processing, and the
detailed instructions to the milling machine, are all solved
in advance. In this way the cobbler simply goes through a
prescribed process in data acquisition and processing,
informing the computer as required about various choices
left for the user (and their clients) to decide.
THE PROJECT
The mensuration section requires several facets
1 To capture the shape of the human foot. Clearly
there is a high predictability about this, when
probably 95% of the adult population's feet can be
sufficiently(?) described by some 15 to 20 lengths
("size" and "half-size"), with two or perhaps three
widths in each. This project is directed at the other
5% of the population, having sufficient variation
