A lexicon for consistent description of material images for nuclear forensics
Explore the source record for details and available documents.
Engineering topics
Publications and source records attributed to Porter, Reid B..
Explore the source record for details and available documents.
This document reviews how the uncertainty in the calculations is being determined with test image data. The results of this testing give an ‘initial uncertainty’ number than can be used to estimate the ‘back end’ uncertainty in digital image quantification in images. Statisticians are refining these numbers as part of a UQ effort.
This document reviews the six ways to segment with the MAMA software. Mama is not a threshold based segmentation system, like many of the image analyses programs you may be have used in the past. The MAMA software is set up to allow you to segment iteratively, and do so in a variety of ways. Six different ways are show below. They can be used in almost any combination—sometimes the choice depends on the image, sometimes it depends on user preferences for what type of interactions a user is willing to do. In the examples below, by “easy”, we mean an image for which MAMA segmentation works great and little clean-up is needed. By hard image, we mean an image for which accurate segmentation is very challenging. Many metal micrographs, and particles with rough textures fall into this class.
This document reviews the sources of uncertainty, standards, and the UQ development plan for the MAMA software. Our intention is to allow users to generate quantification data with uncertainty based on the contributing back end sources. We do not intend to add in front end uncertainty into the software itself, due to the numerous different instruments that could be used, and the enormous variation in factors that would add to that uncertainty among different users, samples, and sample prep routines. We could not possibly or reasonably address them all. We will however, add in the uncertainty component from the image scale bar, which must be provided by the user as an input to the software, and could be used to incorporate any front end uncertainty. (Note: the detailing of these front end uncertainties is beyond the scope of this software effort.) The verification testing we have done generates uncertainties that are fundamentally pixel based. Given the uncertainty in the scale bar, we can then propagate the uncertainty and convert to uncertainty in all calculated results in microns (or whatever unit the scale bar is in.) Users can then add any additional instrumentation and sample prep uncertainty to the resulting output as appropriate for the data calculation and analysis being performed.
This document describes the current state of the MAMA (Morphological Analysis of Materials) software user identified bugs, issues, and requests for improvements. It also lists Current users and current training methods.
This document reviews the verification of the basic shape quantification attributes in the MAMA software against hand calculations in order to show that the calculations are implemented mathematically correctly and give the expected quantification results.
The table on the following pages lists a set of morphology terms for describing materials. We have organized these terms by categories. Software uses are welcome to suggest other terms that are needed to accurately describe materials. This list is intended as a initial starting point to generating a consensus terminology list.
This document reviews the size and shape attributes that are quantified by the MAMA software, shows how they are calculated, and then gives some visual example shapes.
This document shows examples of the results from quantifying objects of certain sizes and types in the software. It is intended to give users a better feel for some of the quantification calculations, and, more importantly, to help users understand the challenges with using a small set of ‘shape’ quantification calculations for objects that can vary widely in shapes and features. We will add more examples to this in the coming year.