Currently, predict does not resize the flatfield/darkfield images and will throw an error if input images are not the working_size.
What should happen is the input image size should be compared to the basic.flatfield array, and then basic.flatfield should be stored as basic._flatfield. Subsequent predict calls should check the inputs images against basic._flatfield and basic.flatfield, and if neither matches then resize basic.flatfield and store it in basic._flatfield. The same applies to basic.darkfield.
There are other ways to do this, but the main goal in this approach is to always keep the original, unmodified flatfield and darkfield data. The original flatfield and darkfield data are what should be saved/loaded from disk.
Currently,
predictdoes not resize the flatfield/darkfield images and will throw an error if input images are not theworking_size.What should happen is the input image size should be compared to the
basic.flatfieldarray, and thenbasic.flatfieldshould be stored asbasic._flatfield. Subsequentpredictcalls should check the inputs images againstbasic._flatfieldandbasic.flatfield, and if neither matches then resizebasic.flatfieldand store it inbasic._flatfield. The same applies tobasic.darkfield.There are other ways to do this, but the main goal in this approach is to always keep the original, unmodified
flatfieldanddarkfielddata. The originalflatfieldanddarkfielddata are what should be saved/loaded from disk.