首页|基于双通道的多维域水声目标识别

基于双通道的多维域水声目标识别

扫码查看
在海洋遥感领域,水声目标分类识别一直是声呐系统的一项困难而又极其重要的任务,为了进一步提高在不同信噪比下水下声目标的识别准确率,本文提出一种使用多维域融合特征分别输入双通道模型的水声目标识别方法.首先,通过梅尔频率倒谱系数(MFCC)和短时傅里叶变换(STFT)提取声信号在频域和时频上的特征进行融合;其次,构建密集卷积神经网络(DenseCNN)和长短期记忆网络(LSTM)2个通道,DenseCNN通道架构采用跳跃连接重用所有以前的特征映射,以优化各种受损条件下的分类率,并采用SE注意力机制使得动态调整特征权重.LSTM通道捕捉时间相关性,对模型进行长依赖关系处理能力的补充.实验结果表明,该方法在-20~10 dB信噪比下的分类准确率优于其他先进的神经网络模型.
Multi-dimensional underwater acoustic target recognition based on dual channel
In the field of Marine remote sensing,the classification and recognition of underwater acoustic targets has al-ways been a difficult and extremely important task for sonar systems.In order to improve the accuracy of underwater acous-tic targets under different signal-to-noise ratios,this paper proposes a method of underwater acoustic target recognition using multi-domain fusion features to input dual channel models respectively.First,the features of acoustic signal in frequency do-main and time frequency are extracted by Mel-Frequency Cepstral Coefficients(MFCC)and short-time Fourier transform(STFT).Secondly,dense convolutional neural network(DenseCNN)and long short term memory network(LSTM)are con-structed.The DenseCNN channel architecture uses skip connections to reuse all previous feature maps to optimize classifica-tion rates under various damaged conditions,and SE attention mechanism enables dynamic adjustment of feature weights.LSTM channels capture temporal dependencies and complement the model's ability to handle long dependencies.Experi-mental results show that the classification accuracy of the proposed method is better than other advanced neural network models at-20~10 db SNR.

underwater acoustic target recognitionmultidimensional time-frequency characteristicsdual channel learning module

张晨颖、杨琼、刘枫

展开 >

西安工程大学计算机科学学院,陕西西安 710600

陕西省服装设计智能化重点实验室,陕西西安 710600

水声目标识别 多维时频特征 双通道学习模块

陕西自然科学青年基金资助项目

2021JQ693

2024

舰船科学技术
中国舰船研究院,中国船舶信息中心

舰船科学技术

CSTPCD北大核心
影响因子:0.373
ISSN:1672-7649
年,卷(期):2024.46(20)