Uncertainty model for automated gunshot residue particle length measurements obtained from electron microscopy images.

Currently, the use of algorithms and computer vision systems for metrological purposes has increased in different areas of knowledge to reduce human error and process deviations, consequently increasing reliability and reducing measurement uncertainties. This study presents a model for estimating the uncertainty of Feret's diameter (DF ) measurements of scanning electron microscopy (SEM) images from regular and irregular gunshot residue (GSR) particles at different magnifications. The data were extracted using the automatic measurement algorithm developed by the Brazilian Institute of Metrology, Quality and Technology (Inmetro). The proposed uncertainty model was based on the recommendations of the guide to the expression of uncertainty in measurement (GUM). The gold standard technique to identify and detect GSR particles is the SEM coupled to energy dispersive X-ray spectroscopy (SEM/EDS), which was used in the study. The low uncertainty values obtained in this study are justified by the refinement of the measurements performed at each stage of digital image procedures. The proposed uncertainty model contributes in an innovative way to the metrological evaluation of regular and irregular GSR particles at different images magnifications. The correct morphometry definition of these particles allows to study their distinction from other possible sources of GSR and, above all, their correlation with the type of ammunition used when firing the firearm. These measurement uncertainty calculations can be applied to any object images acquired by SEM, which provides more confidence in the results of measurements of the object of interest.

Comparison of Chemometric Problems in Food Analysis Using Non-Linear Methods.

Food analysis is a challenging analytical problem, often addressed using sophisticated laboratory methods that produce large data sets. Linear and non-linear multivariate methods can be used to process these types of datasets and to answer questions such as whether product origin is accurately labeled or whether a product is safe to eat. In this review, we present the application of non-linear methods such as artificial neural networks, support vector machines, self-organizing maps, and multi-layer artificial neural networks in the field of chemometrics related to food analysis. We discuss criteria to determine when non-linear methods are better suited for use instead of traditional methods. The principles of algorithms are described, and examples are presented for solving the problems of exploratory analysis, classification, and prediction.