Image data flow
177
Goldeye G/CL Technical Manual V4.1.1
Automatic image control
Definitions
To properly explain the working principle of automatic image control, the
definition of several special terms is clarified first. Refer also to Figure 56 on
page 178.
Accounted pixels
Once the outliers are discarded from the calculation, all pixels holding gray values
in the middle of the image histogram constitute the Accounted pixels.
Mean value
The average gray value of all accounted pixels. This means that within the
accounted pixels range of the image histogram, the same number of accounted
pixels is situated to either side of the mean value.
Minimum, maximum
The smallest and the largest gray value of pixels accepted within the number of
accounted pixels.
Outliers
Outliers are defined as the brightest and darkest pixels in an image. Usually they
originate from defect pixels of the sensor itself.
Small, very bright or very dark objects in the observed scene can also create
outliers. Because very small objects usually do not carry much useful scene
information, they can be treated like outliers as well.
In an image histogram, the outliers show on the far-left side (dark pixels) and on
the far-right side (bright pixels).
Eliminating outliers
The calculations of all auto level control features, including the automatic exposure
operation, may be influenced by outliers in an unwanted way.
The auto level control settings allow for discarding outliers and thus preventing
their undesired impact on the auto level control adjustments.
•
To exclude the outliers on the dark side of the histogram (left), use the
AutoModeOutliersDark
feature. To include all dark pixels into the
calculations, set the feature to
0
.
•
To exclude the outliers on the bright side of the histogram (right), use the
AutoModeOutliersBright
feature. To include all bright pixels into the
calculations, set the feature to
0
.
The values for both exclusion areas do not need to be equal.