February 05, 2010

Auto Channel Balance prevents pattern formation in black-and-white sensors

The camera manufacturer VRmagic has developed an algorithm that automatically corrects pattern formation in black-and-white sensors.


A well-known problem of black-and-white sensors is pattern formation in the gray area. This phenomenon occurs if the same hardware is used for the black-and-white sensor as for a color sensor. Although the sensor surface does not then have a Bayer pattern, which filters the analog color information according to the colors red, green and blue, the signal still passes through three separate channels in the hardware. In color sensors, these three amplifier circuits allow individual adjustment of the colors, e.g. for white balancing. If used with a black-and-white sensor, the separate channels lead to different gain values for neighboring pixels. In order to prevent the resultant patterns in the image, it was necessary in the past to manually adjust the gain. VRmagic has now developed an algorithm that automatically measures and corrects differences in pixel gain. The function has been integrated in the API and CamLab from VRMagic under the name “Auto Channel Balance” and is currently implemented for the camera models VRmC-9/BW, VRmC-9+/BW, VRmFC-9/BW and VRmDC-9/BW.