Recognition of Pen‐Holding Postures: A Child‐Targeted Image Dataset and a Homologous Multimodal Feature Fusion Model
IET Biometrics · 2026-01 · 2026 (1)
DOI: 10.1049/bme2/1263126 ↗Kurz gefasst
Fachartikel in der Zeitschrift IET Biometrics, online erschienen am 15.11.2026. Verfasst von Zheng Yang und 1 weiteren Autor:innen. Die Arbeit nennt 30 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Pen‐holding postures (PHPs) recognition for children is an important topic in computer vision because it has the potential to improve the writing posture of children and prevent them from the risk of developing myopia and other health problems. However, there are no public datasets of the PHPs of children. To fill this gap, 40 primary students are invited to capture their nine different PHPs from the same shooting angle. In addition, a homologous multimodal feature fusion model for PHPs recognition was built on the new dataset in order to mind a more comprehensive essential image feature. The constructed model combines the multidimensional image features with the angle numerical features of mathematical calculation, which is then inputted into the ensemble classifier for prediction purposes. Experimental results of the proposed model showed that it is an effective PHPs recognition model with its benefits of low development costs.
Zitieren
APA
Yang, Z., & Chan, K. K. (2026). Recognition of Pen‐Holding Postures: A Child‐Targeted Image Dataset and a Homologous Multimodal Feature Fusion Model. IET Biometrics, 2026(1). https://doi.org/10.1049/bme2/1263126
Vancouver
Yang Z, Chan KK. Recognition of Pen‐Holding Postures: A Child‐Targeted Image Dataset and a Homologous Multimodal Feature Fusion Model. IET Biometrics. 2026;2026(1). doi:10.1049/bme2/1263126
BibTeX
@article{Yang2026,
author = {Yang, Zheng and Chan, Kan Kan},
title = {Recognition of Pen‐Holding Postures: A Child‐Targeted Image Dataset and a Homologous Multimodal Feature Fusion Model},
journal = {IET Biometrics},
year = {2026},
volume = {2026},
number = {1},
doi = {10.1049/bme2/1263126}
}
EUVIMED