EUVIMEDEUVIMED entdecken
Crossref

A Robust Image Preprocessing Framework for Early Autism Spectrum Disorder Detection Using Multimodal Facial Data

V Visalakshi, D. Maheswari

Dandao Xuebao/Journal of Ballistics · 27.05.2026 · 38 (1) · 553-565

DOI: 10.52783/dxjb.v38.302 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Dandao Xuebao/Journal of Ballistics, online erschienen am 27.05.2026. Verfasst von V Visalakshi und 1 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Early detection of Autism Spectrum Disorder (ASD) remains a significant challenge due to the variability in behavioral manifestations and limitations of traditional diagnostic approaches. Although machine learning techniques have shown promising results in automated ASD detection, their effectiveness largely depends on the quality and consistency of input data. This study proposes a robust and scalable image preprocessing framework specifically designed for multimodal facial datasets to enhance feature representation and improve classification readiness. The framework integrates adaptive median filtering for noise reduction, histogram equalization for contrast enhancement, facial alignment for geometric normalization, data augmentation to improve generalization, and HSV color space transformation for enhanced feature separability. Additionally, region-of-interest (ROI) extraction is employed to isolate salient facial regions relevant for analysis. The proposed framework is evaluated on a dataset comprising 2,928 facial images, demonstrating notable improvements in data quality and subsequent classification performance. The findings emphasize the critical role of preprocessing in developing reliable and efficient AI-driven diagnostic systems for early ASD detection and highlight its potential for integration into real-world engineering-based healthcare applications.

Zitieren

APA

Visalakshi, V., & Maheswari, D. (2026). A Robust Image Preprocessing Framework for Early Autism Spectrum Disorder Detection Using Multimodal Facial Data. Dandao Xuebao/Journal of Ballistics, 38(1), 553-565. https://doi.org/10.52783/dxjb.v38.302

Vancouver

Visalakshi V, Maheswari D. A Robust Image Preprocessing Framework for Early Autism Spectrum Disorder Detection Using Multimodal Facial Data. Dandao Xuebao/Journal of Ballistics. 2026;38(1):553-565. doi:10.52783/dxjb.v38.302

BibTeX
@article{Visalakshi2026,
  author = {Visalakshi, V and Maheswari, D.},
  title = {A Robust Image Preprocessing Framework for Early Autism Spectrum Disorder Detection Using Multimodal Facial Data},
  journal = {Dandao Xuebao/Journal of Ballistics},
  year = {2026},
  volume = {38},
  number = {1},
  pages = {553--565},
  doi = {10.52783/dxjb.v38.302}
}
Kostenlos starten →