Measurement, modeling and visualization
of the shape of the cornea
F.M. Vos et al.
[Abstract],
[Introduction],
[Image Processing],
[Surface Reconstruction],
[Conclusions and Future Work],
[References]
Section 4: Conclusions and future work
Our newly developed instrument exploits the properties of Pseudo
Random Binary Arrays to capture the corneal surface structure. Encoded
in a colored stimulus pattern PRBA's allow for a unique characterization
of positions.
Given these data splines can conveniently be used to model the surface
measured. Experiments with well defined test objects show that the methodology
thus defined is extremely accurate.
Critical to this application is the color determination and graph recovery.
The procedure outlined in this paper provides an excellent starting point
for both. It can be expected that better results can be achieved when the
procedure is iteratively applied.
The modeling is at best performed in a highly interactive way. Currently,
even using state of the art uniprocessors, this goal can not be achieved
for high numbers of patches. However the problem is characterized by a high
intrinsic parallelism. Effort is therefore devoted to parallelize this problem
thus achieving the desired speed up.
For the presentation and interpretation of the experimental results at all
stages visualization is indispensible. We find DX an attractive environment
to program and develop prototype user interfaces.
[Abstract],
[Introduction],
[Image Processing],
[Surface Reconstruction],
[Conclusions and Future Work],
[References]