ISAR Dataset for the Recognition of Conical Targets with Micro-Motion

Abstract In recent years, the recognition of ballistic micro-motion targets based on deep learning has been extensively studied. However, currently, there are no publicly available datasets; all datasets come from simulations conducted by researchers themselves. In this study, it was found that even...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhichen Zhao, Degui Yang, Xing Wang, Jianxuan Xu
Format: Article
Language:English
Published: Nature Portfolio 2025-06-01
Series:Scientific Data
Online Access:https://doi.org/10.1038/s41597-025-05193-4
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Abstract In recent years, the recognition of ballistic micro-motion targets based on deep learning has been extensively studied. However, currently, there are no publicly available datasets; all datasets come from simulations conducted by researchers themselves. In this study, it was found that even when the motion parameters and model are kept the same, the details of the electromagnetic simulation method have a significant impact on the data. Therefore, there is an urgent need for a publicly available dataset to evaluate the performance of different methods. The ISAR Micro-Motion Dataset (IMD) is a simulated radar echo dataset based on the working principles of fully polarimetric ISAR. It consists of two components: aspect angle sequence data and static electric field data of the target. This paper presents a unified process for generating target radar echoes and discusses how various details can impact the results.
ISSN:2052-4463