What is it about?
We investigate whether AI can distinguish between passenger and military aircraft using images, particularly when traditional identification signals may be unavailable. More than 15,000 aircraft images were used, including images with different levels of noise, lighting conditions, and viewing angles. Three AI approaches were compared, and the capsule network achieved the best performance, with about 97% accuracy.
Featured Image
Photo by Bing Hui Yau on Unsplash
Why is it important?
Misidentifying aircraft is a serious safety concern in shared airspace. This study explores whether image-based classification could one day support existing surveillance methods. It provides initial comparative results under simulated conditions.
Perspectives
This work may be of interest to researchers working on aircraft recognition, remote sensing, and capsule networks. The results also highlight open questions, including testing with real operational imagery, additional sensor types, and reducing computational cost.
Zainab Malik
Read the Original
This page is a summary of: From radio waves to pixels: deep learning for aircraft identification systems, Aerospace Systems, October 2025, Springer Science + Business Media,
DOI: 10.1007/s42401-025-00414-5.
You can read the full text:
Contributors
The following have contributed to this page
