Adding Slap Image Deficiencies

[Slap Fingerprint Segmentation (SlapSeg)]

Thanks for your interest in SlapSeg III! We've made a few changes to the test plan and API over the last couple of days that we wanted to bring to your attention. Most notably is the concept of image deficiencies.

SlapSeg III will be an offline test, meaning that algorithms will be presented with slap images where the subject is not present. However, we still want to test whether the algorithm can reason if, and why, a recapture of the slap image should be required when presented with a given image, pretending as if the subject was present. We've added a list of image deficiencies to choose from in these scenarios. Set one or more of these in the ReturnStatus object when suggesting a recapture. We'd appreciate your feedback and suggestions for inclusion in this list.

Failure to segment should be a last resort, and a best-effort segmentation should always be provided if possible. For instance, if it is determined that an image shows "ghost" low-contrast impressions from a previous capture, a robust algorithm can likely still succeed at segmentation, while specifying in the ReturnStatus that the platen was dirty. However, if the hand is placed sideways or upside down, it may be unwise to attempt the segmentation.

Thanks again for your continued feedback.