G-7
G
Glossary
Ë
Island
An "island" is a separate area which is created after labeling (object identification) process the binary
image.
[I]
Ë
Illumination monitor alarm density setting
The "illumination (light level) monitor" is a function which automatically monitors the environmental light-
ing conditions when measuring objects.
If the illuminance exceeds the alarm density setting, the IV-S20 will display an alarm message.
Ë
Interline transfer system
The interline transfer system is an electrical charge transfer system which consists of two separate ar-
eas: the area where the light beams are converted into electrical charges by the CCD elements, and the
area where the charges are transferred.
+
30
°
−
30
°
[M]
Ë
Main axis angle
The "center of gravity" is the geographical center of the image. It is determined by treating the binary
image to be measured as an object that has mass.
Ë
Mask window
The image being examined may contain an area that does not need to be processed. To eliminate such
an area, a mask window is used.
Measurement window
Object to be measured
Mask window
[N]
Ë
Normalization correlation
To determine whether the levels between the reference image and the input image match, the IV-S20
uses an information processing method called "normalization correlation." This is a method used to cal-
culate relationship between two groups of data.
- Factors determining the correlation value
If the densities of the two images have the same tendency (positive correlation), the two images are
said to resemble each other. If the densities of the two images have opposite tendencies (negative
correlation), the two images are said not to resemble each other. Therefore, areas of the reference
image and the input image which resemble each other (the areas of both images are brighter, or
darker) are positive, and areas which do not (the areas of one is brighter and the other is darker) are
negative.
- Correlation formula = { A
√
B x C } x 10000
A = N
∑
(I x T) – (
∑
I) x (
∑
T): Correlation between input image and reference image
B = N
∑
(I x I) – (
∑
I) x (
∑
I): Correlation between input images
C = N
∑
(T x T) – (
∑
T) x (
∑
T): Correlation between reference images
(N: Area of reference image, T: Density of reference image, I: Density of input image)