2.2 Data processing
2.2.1 Computer hardware: A Research Machines PC-560
Professional, equipped with a 60 MHz Pentium processor,
32 Mb RAM, 1Gb hard disk, 2Mb graphics accelerator card
and a 17 inch Super Video Graphics Array (SVGA) monitor
was employed in this project. Such a machine, which delivers
workstation-like performance, is necessary for the efficient
handling of large, digital data files and currently costs around
$3000.
2.2.2 Computer software: Several software packages were
used in this project.
Adobe Photoshop v3.0.5, which is an advanced image
processing and photo-retouching package, was used as an
electronic darkroom. This package will accept imagery in
colour or grey scale, obtained from digital cameras and by
scanning conventional photography, in a variety of formats.
The current cost of this package is $700.
For measurement, the R-Wel Inc. Desktop Mapping System
(DMS) v4.0 was selected. This is a PC-based, low cost
($6000), digital photogrammetric system incorporating image
processing operations, automatic image correlation and
stereoplotting capabilities. Further details of the system can be
found in Welch (1989).
The differences between the back measurements produced
digitally and analytically were analysed with the Land Survey
System (LSS) software package produced by McCarthy Taylor
Systems Ltd. This is a PC-based DTM and mapping package
currently costing $3500.
The total cost of the hardware and software used is in the order
of $25000 at current prices. This is low in terms of the general
high cost of photogrammetric equipment.
3. APPLICATION 1 - PLANNING OF ORAL SURGERY
This application, undertaken in conjunction with the University
School of Dentistry, relates to the use of digital imagery in the
planning of oral surgery for the correction of facial deformity.
This is an application where conventional imagery is already in
use (Fanibunda, 1983) and in which it is necessary to
superimpose several lateral view images of the patient's head
(in particular, a conventional profile image and a skull X-ray)
in order to get the full picture for planning purposes. The
resulting scaled composite image, which portrays the hard and
soft tissues of the face in their correct relationship, is of great
help to the surgeon in planning an appropriate operative
procedure for the patient. The particular planning that this is
designed to help is in relation to surgical repositioning of the
upper or lower jaw, and the effects of this movement on the
patient's facial profile. The final profile is obviously of some
concern as the surgical procedure is of a permanent nature.
Consequently planning must be carried out as accurately as
406
possible and anything that can be done to put this on a more
scientific footing is highly desirable. A positioning accuracy in
the order of +/- 1 mm is required.
The composite image is currently produced, with some
considerable difficulty, through normal photographic darkroom
techniques. The use of digital imagery has the potential to
make the handling of these images and their superimposition
much easier. Further, the planning of the surgery, which
typically involves a movement of the jaws, can be undertaken
by simply cutting and moving a portion of the final digital
image on the monitor screen and several different solutions can
be quickly assessed.
3.1 Method
The imagery of each patient was taken in the School of
Dentistry, where radiological equipment is available which
could be calibrated for use in this work.
When taking the imagery, the patient's head was immobilised
in a cephalostat, as employed for routine radiographs, by means
of two ear-rods. A projected line of light and adhesive targets
placed on the skin were used to ensure that the head was kept
in a constant orientation, with the Frankfurt plane horizontal.
A 300 mm steel ruler with serrated edges was attached to the
cephalostat in the mid-sagittal plane and aligned with its edges
vertical. This ruler, being steel, imaged on the radiograph as
well as the digital image.
The cephalostat was permanently mounted on a wall and the X-
ray tube could be positioned such that its principal axis was
perpendicular to the mid-sagittal plane and the principal point
coincided with the centre of the ear-rods of the cephalostat.
Further, its focal spot was positioned at a fixed distance
(approx. 2 m) away from the radiographic film plane. When in
this position a radiograph was exposed.
The X-ray tube was then moved vertically upwards by a fixed
amount. A mount (Figure 1), fixed to the X-ray tube, allowed
the Kodak DCS200 camera to be positioned such that the front
nodal point of its lens coincided with the focal spot of the X-ray
tube when the radiograph was taken. The principal axis of the
camera in this position was again perpendicular to the mid-
sagittal plane and the principal point coincided with the centre
of the ear-rods. After positioning the camera, an exposure Was
made.
The digital colour image from the Kodak DCS200 camera,
originally stored in a Kodak proprietary format, Was
downloaded from the camera's hard disk through a SCSI
controller using the Kodak Twain driver into Adobe Photoshop.
Various image enhancement techniques were then applied to
improve, as necessary, such features as brightness, contrast and
colour balance.
After development, the radiograph was in the form of à
diapositive. This was then scanned at 300 dpi using the
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B5. Vienna 1996
Figure 1 -
HP ScanJet
the resultin
brightness,
and subseq
in Adobe F
because thi
even after )
at the tim
Consequen
instead of
image. TI
camera wa:
light table
ear-rods. *
radiograph
better ime
Photoshop,
to highligh
The two in
to a comm
photograph
two serrat;
measured |
image. |
