Banca de QUALIFICAÇÃO: CARLOS MONTEIRO DE LIMA

Uma banca de QUALIFICAÇÃO de DOUTORADO foi cadastrada pelo programa.
STUDENT : CARLOS MONTEIRO DE LIMA
DATE: 28/08/2025
TIME: 14:00
LOCAL: videoconferência
TITLE:

Development of methodologies and instruments for beer authentication using Chemometrics and digital images.

KEY WORDS:

Beer authentication; Digital images; Chemometrics; DD-SIMCA; AFTM-Led device

PAGES: 102
BIG AREA: Ciências Exatas e da Terra
AREA: Química
SUBÁREA: Química Analítica
SPECIALTY: Instrumentação Analítica
SUMMARY:

This study proposes an approach for the authentication of commercial beers based on histograms extracted from digital images acquired using a prototype of the Fluorimeter and Turbidimeter Device with LED Excitation and Digital Image-Based Detection (AFTM-Led), developed by the authors. From the acquired images, a region of interest (ROI) was selected and used for sampling and quantization, generating histograms in the RGB (Red-Green-Blue), HSI (Hue-Saturation-Intensity), and grayscale (GS) color channels. A total of 30 beer samples—comprising both commercial and homemade varieties—were analyzed and categorized into target and non-target classes. Chemometric analysis was carried out using principal component analysis (PCA) and one-class modeling via DD-SIMCA (Data-Driven Soft Independent Modeling of Class Analogies). Model validation was performed using a pseudo-validation set generated through the Procrustes cross-validation (PCV) approach. PCA revealed clear clustering of the target class, although some overlap with non-target samples was observed. Among the DD-SIMCA models, the one based on GS channel data demonstrated high sensitivity and specificity, achieving 100% accuracy in both modeling and non-target class prediction, with only two errors observed during PCV validation, resulting in an overall efficiency of 93.4%. The use of the AFTM-Led device confirmed the practical feasibility of the proposed method, standing out as a compact and accessible instrumental solution.

COMMITTEE MEMBERS:
Presidente - 1714946 - KASSIO MICHELL GOMES DE LIMA
Interna - 3420591 - ANA CAROLINA DE OLIVEIRA NEVES MENEZES
Externa à Instituição - ANNE BEATRIZ FIGUEIRA CÂMARA - UFRN
Externo à Instituição - CAMILO DE LELIS MEDEIROS DE MORAIS - UECE