SVS-Vistek 926.15hr387*XGE Series, Руководство пользователя

Страница: 35 / 122

SVS-VISTEK
Feature description 35
5.2.6 Defect Pixel Correction
All image sensor have defect pixels in a lesser or greater extent. Type and
number of defects determine the quality grade (quality classification) of
the sensor.
Defect Pixel Correction is using information from neighboring pixels to
compensate for defect pixels or defect pixel clusters (cluster may have up
to five defect pixels).
Defect Pixels either be dark pixels, i.e. that don’t collect any light, or
bright pixels (hot pixel) that always are outputting a bright signal.
The amount of hot pixels is proportional to exposure time and
temperature of the sensor.
By default, all known defect pixels or clusters are corrected by SVS-VISTEK
as a factory default.
Under challenging conditions or high temperature environments defect
pixel behaviour might change. This can be corrected.
>
A factory created defect map (SVS map), defying known defects, is
stored in the camera.
>
A custom defect map can be created by the user. A simple txt file with
coordinates has to be created. The user must locate the pixel defects
manually.
>
The txt file can be uploaded into the camera. Beware of possible
Offset!
>
Defect maps can be switched off to show all default defects, and
switched back on to improve image quality.
Unlike Shading Correction, Defect Pixel Correction suppresses single
pixels or clusters and reconstructs the expected value by interpolating
neighboring pixel values. The standard interpolation algorithm uses the
pixel to the left or to the right of the defect. This simple algorithm prevents
high runtime losses.
More sophisticated algorithms can be used by software.
Figure 3: Illustration of a defect pixel