Skip to content

Confidence Threshold#

The Confidence Threshold feature allows you to exclude unreliable data from your measurements with blaze cameras.

Using the Feature#

How It Works#

The camera calculates an internal confidence value between 0 and 65335 for every pixel. The more light is collected for a pixel, the higher the confidence value. With the ConfidenceThreshold parameter you can specify that only pixels above a certain confidence value are included in the camera output. If the internal confidence value of a pixel doesn't pass the confidence threshold, the corresponding pixel position is represented in the various components like outlined in the following table:

Component Representation of Pixels Below the Confidence Threshold
Range For point clouds, i.e., the PixelFormat is set to Coord3D_ABC32f, all 3 coordinates of a pixel are set to the value specified by the Scan3dInvalidDataValue parameter.
For depth maps, i.e., the PixelFormat is set to Coord3D_C16 or Mono16, the depth value for a pixel is set to the value specified by the Scan3dInvalidDataValue parameter.
Intensity The confidence threshold has no effect on the Intensity component. There will always be intensity values for every pixel regardless of the confidence value.
Confidence Set to zero.

Setting a Confidence Threshold#

You can specify a confidence threshold value between 0 and 65335. As conditions can vary a lot in different applications, there is no general guideline for choosing a value. Try experimenting with this parameter to see which setting delivers the best results, i.e., at which setting undesired parts of the image are removed.