Long-Huang Tsai

Mr. Long-Huang Tsai


Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
13 January 2026 in Sensors
Verified
Journal Article
Open access
PDF file
HLHsiang-Ju Lai
Hsiang-Ju Lai
LTLong-Huang Tsai
Long-Huang Tsai
HKHSU KUNGYANG
HSU KUNGYANG
WYWen-Chao Yang
Wen-Chao Yang
With the rapid advancement of technology in recent years, signatures on contracts and documents have increasingly shifted from traditional handwritten forms on paper to digital handwritten signatures executed on devices (hereafter referred to as digital tablets). This transition introduces new challenges for forensic document examination due to the differences in writing instruments. According to the European Network of Forensic Science Institutes (ENFSI), a Digital Capture Signature (DCS) refers to data points captured during the writing process on digital devices such as tablets, smartphones, or signature pads. In addition to retaining the visual image of the signature, DCS provides more information previously unavailable, including pen pressure, stroke order, and writing speed. These features possess potential forensic value and warrant further study and evaluation. This study employs three devices—Samsung Galaxy Tab S10, Apple iPad Pro, and Apple iPad Mini—together with their respective styluses as experimental tools. Using custom-developed handwriting capture software for both Android and iOS platforms, we simulated signature-writing scenarios common in the financial and insurance industries. Thirty participants were asked to provide samples of horizontal Chinese, English, and number writings (FUJ-IRB NO: C113187), which were subsequently normalized and segmented into characters. For analysis, we adopted distance-based time-series alignment algorithms (FastDTW and SC-DTW) to match writing data across different instances (intra- and inter-writer). The accumulated distances between corresponding data points, such as coordinates and pressure, were used to assess handwriting stability and to study the differences between same-writer and different-writer samples. The findings indicate that preprocessing through character centroid alignment, followed by the analysis, substantially reduces the average accumulated distance of handwriting. This procedure quantifies the stability of an individual’s handwriting and enables differentiation between same-writer and different-writer scenarios based on the distribution of DCS distances. Furthermore, the use of styluses provides more precise distinctions between same- and different-writer samples compared with direct finger-based writing. In the context of rapid advancements in artificial intelligence and emerging technologies, this preliminary study aims to contribute foundational insights into the forensic application of digital signature examination.

1 September 2026 in Applied Sciences
Verified
Journal Article
Open access
PDF file
LTLong-Huang Tsai
Long-Huang Tsai
HLHsiang-Ju Lai
Hsiang-Ju Lai
WYWen-Chao Yang
Wen-Chao Yang
JJJIAJUN JIANG
JIAJUN JIANG
CCChung-Hao Chen
Chung-Hao Chen
Digitally captured handwriting preserves pen trajectories and dynamic signals, but it also records hardware- and input-dependent properties that can confound forensic interpretation. This study revises a support vector machine (SVM) screening framework using 16,500 samples from 30 writers, 11 writing-content categories, and five acquisition conditions spanning three tablets and stylus or finger input. Twenty-four raw and derived time-series variables were summarized by maximum, minimum, mean, median, and standard deviation, yielding 120 features; the mode statistic was removed. Writing direction and angular velocity were recalculated with atan2-based vector formulas. Unavailable device/API channels were encoded as zero, and Z-score parameters were estimated only from training folds. Writer and content evaluations used rotating pooled-“other” categories as rejection-class proxies, whereas acquisition-condition classification remained closed-set. Every outer five-fold split contained an inner five-fold forward-selection loop; RBF-SVM hyperparameters were fixed a priori (C = 1.0, gamma = scale, balanced class weights, and random seed 42). Writer classification achieved 94.85% accuracy (descriptive 95% CI: 94.69–95.01%) and 86.44% pooled-other recall. Content classification achieved 96.64% accuracy (95% CI: 95.74–97.53%) and 98.74% pooled-other recall. Acquisition-condition classification achieved 99.99% accuracy (99.98–100.00%), with one error among 16,500 outer-test predictions. The acquisition result is interpreted primarily as evidence that channel availability and device-specific measurement scales are strongly encoded in the feature space. Because the folds were sample-level, the samples were collected contemporaneously, and pooled-other writers were represented during training, these results do not establish session-disjoint, cross-device writer, or strict open-set generalization. The proposed workflow should therefore be regarded as an experimental triage aid that supports, rather than replaces, examiner-led comparison.