v3.2.3 Release Notes

New Functions

Crop Image > Apply Last crops the image using the area defined in the last Crop Image step in the recipe.

Fixed bug with global measurement report printing
Fixed bug with applying pre-loaded model from Deep Learning Trainer
Fixed bug where Deep Learning Trainer did not respect GPU selection preference
Fixed bug with certain workflows for flagging recipe steps and setting notes
Fixed bug where setting scale factor could cause errors during recipe update when first step was Crop Image
Fixed bug with adding new layers after previous layers were cleared
Fixed bug where error messages could appear behind Apply Model window
Fixed bug where Pattern Mapping would not run on a binary image during recipe update

In Reject Features “Proximity Awareness”, Local Window settings now include the feature potentially being rejected
Faster preference reading
Removing all recipe steps now behaves identical to pressing “Clear” (trash can)
Future auto-updates will now use AWS-hosted links
Posted workaround for adding custom logos into report templates

Individual feature measurements are now output
Images can now be input as numpy arrays in addition to filenames
Load Companion images can now be input separately, allowing for more powerful batch processing solutions
Installer has been streamlined
Error and warning reporting has been improved