重庆理工大学学报2025，Vol.39Issue(9)：131-138,8.DOI:10.3969/j.issn.1674-8425(z).2025.05.016
深度学习驱动的无人机噪声指纹检测与识别
Deep learning based drone detection via noise fingerprinting
刘宇畅 1刘涛 2孙宗林 1冯宇 1魏斯洋2
作者信息
- 1. 西南电子设备研究所,成都 610036
- 2. 中国人民解放军93142 部队,成都 610041
- 折叠
摘要
Abstract
A detection and identification scheme using noise fingerprints in drone flight is proposed,which achieves low cost,little environmental impact and fast operation.In the data pre-processing stage,resampling is employed to enhance the data,reduce the impact of environmental noise,and improve the adaptability of the system in different environments.The feature extraction module lessens the impact of signal strength and distance changes on feature stability through the time domain peak normalization algorithm and the feature vector rescaling algorithm.It also enhances the robustness of the system.In the drone detection and identification stage,a CNN and LSTMhybrid deep learning model is constructed.A voting mechanism is introduced to consider the local and overall characteristics of the audio signal,thereby improving the recognition accuracy.Experimental results show the system accurately detects and identifies drones with the presence of different environmental noises and distance differences.关键词
声学特征/声音身份识别/无人机检测/无人机识别/深度学习Key words
acoustic features/voice identification/drone detection/drone identification/deep learning分类
计算机与自动化引用本文复制引用
刘宇畅,刘涛,孙宗林,冯宇,魏斯洋..深度学习驱动的无人机噪声指纹检测与识别[J].重庆理工大学学报,2025,39(9):131-138,8.
