user manual

6.4.3
6 - 63
(4) Baseline Detection
The baseline is a straight line having no slope, which is determined by the peak position
attained before, the "baseline start point" and "baseline area".
The start point of the baseline is located outside or inside the peak position, and it
corresponds to a position of data having differential coefficient "dbase" in the following
expression of profile differentiation.
d
base
= .............................................................. (6.4.4)
d
peak
: Differentiation coefficient of peak position
B : Baseline start point (1, 2, 4, 8)
To determine the baseline, averaging is performed on a profile area included in a range
specified with "baseline area". For the "baseline start point", it is allowed to enter any one
of 1, 2, 4 and 8.
(5) Length Detection
For length detection, an intersecting point of slope line and baseline is determined.
Then an actual pattern dimension is calculated using the intersecting point data,
magnification and correction coefficient.
6.4.3.3 Threshold Level Method
Described next is the threshold level algorithm. In the same manner as for the linear
approximation method, a smoothed profile "s (n)" is obtained. Maximum and minimum values
of the profile are determined in the edge detection area. Then an intersecting point with "Th"
given by the following expression is defined as the edge position.
Th = × T + s
min
.............................................. (6.4.5)
s
max
: Max. value of profile in the vicinity of edge
s
min
: Min. value of profile in the vicinity of edge
T : Threshold (%)
d
peak
|B|
s
max
- s
min
100