数据恶劣条件下的辐射源个体识别方法综述

Overview of individual identification methods for radiation sources under harsh data conditions

摘要:

本文主要针对数据恶劣条件下的辐射源个体识别方法进行了分析对比。总结了包括不平衡,错标签、小样本和弱标注四种情况下的个体识别方法,探讨辐射源特征提取方法的优点和局限性,对方法中作为技术关键和难点的特征提取方法也进行了概括,并指出深度学习在深度特征提取上的优势以及在辐射源个体识别领域的广泛应用前景,以期对各种情况下的辐射源个体识别方法做出较为全面的补充。

Abstract:

This article mainly analyzes and compares individual identification methods for radiation sources under harsh data conditions. Summarized individual recognition methods including imbalance, mislabeling, small samples, and weak labeling, explored the advantages and limitations of radiation source feature extraction methods, summarized the key and difficult feature extraction methods in the methods, and pointed out the advantages of deep learning in deep feature extraction and its broad application prospects in the field of radiation source individual recognition, To provide a comprehensive supplement to individual identification methods for radiation sources in various situations.