Matrizes aumentadas em quimiometria: aplicação na datação de tintas de caneta gel a partir de imagens digitais

Abstract

The dating of inks in questioned documents remains one of the major analytical challenges in forensic document examination due to the complex physicochemical processes involved in ink aging. In this study, the potential of augmented matrices in chemometric analysis was investigated to evaluate the aging behavior of blue and black gel pen inks using digital images. Eighteen gel pen samples from three commercial brands (Cis Spiro, Pilot G2, and Pentel Energel X) were deposited on white paper and monitored over a period of 90 days. Digital images were acquired under controlled conditions, and RGB histograms were extracted to characterize the chromatic profiles of the inks. The resulting data were organized into augmented matrices and analyzed by Principal Component Analysis (PCA). A total of eight principal components were required to explain 91.26% of the total variance, while the third principal component enabled the discrimination between blue and black inks and revealed temporal differences associated with ink aging. The blue channel showed the greatest contribution to sample differentiation. The proposed strategy represents a rapid, non-destructive, and low-cost alternative for forensic ink analysis and document dating applications.