Closed imagingbook closed 1 year ago
I agree this is not intuitive but I can't switch to using Double.NaN without breaking existing plugins.
The ImageJ 1.54a25 daily build adds an ImageProcessor.isThreshold() method. There was already an ImagePlus.isThreshold() method.
It appears that this constant in
ImageProcessor
is initialized with a rather strange value:public static final double NO_THRESHOLD = -808080.0;
By default, this value is returned by both
ImageProcessor.getMinThreshold()
andgetMaxThreshold()
, which does not seem intuitive. How about usingDouble.NaN
instead?