Module: Background Detection Correction ()
A module to normalize volumetric grey-level images by computing an approximation to the background and subtracting it from the image. The image is scaled later to the original input data range.The background image can be computed in different ways dependent on the actual data. The minimum option computes the background by using the minimum value in a region of interest defined by the Filter size port. This mode is especially useful for images with non-uniform illumination and a larger number of objects. The maximum and the mean mode compute the corresponding maximum and mean of minimum and maximum operation. The maximum is, therefore, useful if the background contains larger values than the objects.
The B-Spline mode uses bi-cubic splines to first down-sample the image to the resolution given in the Filter size port. This is usually giving a smoother approximation of the background. The modules minimum, maximum, mean, and B-Spline use a bi-cubic spline to up-sample the low-resolution version back to the original image size before subtracting the background from the data thus correcting for non-uniform illumination.
The Whitening mode aims to normalize the data to zero mean and unit variance. This will set the contrast to be close to uniform throughout the image. You should provide the data as either float or double in this case.
Press the Apply button to start the computation.
Data [required]
The grey-level data set the correction is performed on.
Type
Select a mode for the normalization. Different modes make sense for different images.Filter Size
This is either the resolution of the down-sampled image (for minimum, mean, maximum, and B-Spline) or the size of the moving filter that is used for the whitening operation.Options
You can export the calculated background image as well.