| |
CImage:FiltBlockAvg
The FiltBlockAvg method averages neighboring pixels in blocks of m x n to make a smaller image having better signal to noise ratio in each pixel.
bResult = CImage:FiltBlockAvg( ColBlock, RowBlock, datatype=nil )
ColBlock is the number, m, of adjacent columns to average.
RowBlock is the number, n, of adjacent rows to average.
datatype is an optional data type for the output image, defaults to "same" if not specified.
On success, this method returns true.
On failure, this method returns false.
This method increases the signal to noise ratio by sqrt(mn) for using blocking factors ofm and n. This approximates the effect of using a binning mode during image acquisition except that each blocked pixel will have m x n units of readout noise in its total noise budget whereas a binned pixel of the same blocking factor will have 1 unit of readout noise in its total noise budget. If the unblocked image has an integer type and the output type is also an integer type, then 0.29 counts of quantization noise is also added to each super-pixel.
It may be desirable to change the output data type to a real number variety such as "float" or "double" to accommodate fractional pixel values. However, the gain in statistical significance is usually small. If the current data type is acceptable, then do not specify the optionaldatatype argument.
Suppose an image is displayed in a CImageView named V. The following script fragment block averages the image to 2x2 super pixels and then redisplays the new image:
|
I = V:GetImage() |
-- attach the current displayed image |
|
I:FileBlockAvg(2,2,"float") |
-- average the pixels and convert to 32 bit float. |
|
V:Update() |
-- update the display after the scale |
|
V:SetZoom(2) |
-- magnify the image 2x |