-
...needed
-
Accurate, dense correspondence can be more efficiently determined later
by exploiting the epipolar constraint.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
-
...iterations
-
Of course, numerical implementations of the SVD do actually require
inner iterations.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
-
...similarity
-
If only the correlation information was used (i.e. 39#39 the results
would be extremely poor, as shown in Figure
1
-right, producing a curious ``maximal sum of correlation'' mapping,
which makes clearly no physical sense.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
-
...of
-
The overhead of feature detection and correlation computation are common
to most feature-based matching approaches.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .