5. –
™ Image contains a lot of extra information (jaw, nose,
etc.)
➠ crop to a region of interest (ROI)
™ Characteristics of the radiographs:
– Variation in scale and position of head and teeth is
limited
– Head is always centered horizontally
™ Define the ROI based on the mean of the Gaussian
distribution of the ROI of each image
Region of interest (ROI)
7. –
™ Teeth have a higher grey level intensity than jaws
and other (soft) tissue, because of their higher tissue
density
➠ gap between jaws forms a valley in the y-axis
projection histogram
™ How?
– Determine a set of points that have minimal intensity
– Use interpolation to estimate the gap valley
– Determine a split line (parallel to the x-axis)
Separation of jaws
16. –
™ Segment image based on similar intensity
™ No proper segmentation : toothwas not sufficiently
delineated
➠exterior is also flooded
Water shedding
17. –
= (Partial) solution to the previous problem:
– Remove part of the noise (upper part of the image)
Mean shift filtering
18. –
= (Partial) solution to the previous problem:
– Remove part of the noise (upper part of the image)
– Apply a Gaussian blurring (again)
Mean shift filtering
19. –
= (Partial) solution to the previous problem:
– Remove part of the noise (upper part of the image)
– Apply a Gaussian blurring (again)
– Apply mean shift filtering to smoothen the image
Mean shift filtering
20. –
= (Partial) solution to the previous problem:
– Remove part of the noise (upper part of the image)
– Apply a Gaussian blurring (again)
– Apply mean shift filtering to smoothen the image
– Reapply the water shedding algorithm (and adjust the
image)
Mean shift filtering
22. –
™ Both methods do not work really well
™ The delineation is not good enough to perform a
good classification
– Not implemented
™ Possible methods to consider:
– Hamming distance
– Eigenfaces
– Principal Component Analysis (PCA)
Results from
segmentation
23. –
™ Use segmented image as a mask to compare with the
retrieved segmentation
= compare
™ Determine common scale
– E.g. Smallest box arround both teeth
™ Determine number of not-matching pixels
Hamming distance
24. –
™ Created bitmap that contains most characteristics of
an incisor
™ If a segmented tooth can be described as a weighted
sum of a number of the bitmap images, it is classified
as an incisor
Eigenfaces
25. –
™ Determine the principal components in both the
segmented image and the retrieved segmentation
™ Determine the distance between the principal
components
Principal Component
Analysis (PCA)