Computational Analysis of Dimensional Stability in Reproduced Signatures Using Fiji (ImageJ): A Multi-Device, Multi-Generation Study and Its Forensic Implications

Abstract

The present study investigates the effect of successive reproduction on the dimensional characteristics of handwritten signature samples using a computational image analysis approach. Samples were reproduced up to five generations (G1-G5) using laser multifunction printers (MFPs), inkjet MFPs and single-function photocopiers, and were compared with original control specimens. Dimensional measurements, specifically height and width, were extracted using Fiji (ImageJ), an open-source digital image processing platform, to support precision and reproducibility in measurement acquisition. Statistical analysis was performed using one-way analysis of variance (ANOVA), Dunnett's post hoc test and correlation analysis, with significance set at p <= 0.05. The results indicated that no statistically significant differences were observed in either height or width measurements across all devices and generations (p > 0.05). Minor fluctuations in variability were observed in later generations; however, these remained statistically non-significant. Correlation analysis further revealed no meaningful relationship between height and width variations, indicating independent and random dimensional changes. The findings demonstrate that computational measurement using Fiji provides reliable dimensional extraction, while modern reproduction technologies preserve geometric features with high fidelity. From a forensic and computational perspective, dimensional parameters are stable but possess limited discriminative power for detecting reproduction or forgery. The study highlights the importance of integrating digital image analysis tools into forensic workflows and recommends future incorporation of pixel-level and feature-based computational methods for enhanced document examination.