Page 140 - 《应用声学》2022年第3期

第 41 卷 第 3 期 Vol. 41, No. 3
2022 年 5 月 Journal of Applied Acoustics May, 2022

⋄ 研究报告 ⋄

基于改进MobilenetV2网络的声光图像融合水下

目标分类方法 ∗

巩文静 1,2,3 田 杰 1,3 李宝奇 1,3 刘纪元 1,3†

(1 中国科学院声学研究所 北京 100190)
(2 中国科学院大学 北京 100049)
(3 中国科学院先进水下信息技术重点实验室 北京 100190)
摘要：针对小样本条件下水下目标分类准确率低、计算资源量大的问题，提出一种声光图像融合目标分类
方法。首先，对 MobilenetV2 网络进行改进，去掉第 9 层之后的网络层，并将该层卷积通道数改为 128，通过
Flatten 层进行数据降维，增加一个全连接层得到分类结果；其次，设计一种融合网络结构，将声光图像成对输
入网络进行特征提取，在中间层利用通道拼接算法实现特征图融合，使用融合特征进行目标分类。在真实数据
集上对网络进行训练，结果表明，改进的 MobilenetV2 网络对水下目标的分类性能更好，融合网络的分类准确
率相比融合前有所提高，更加适用于水下目标分类任务。
关键词：改进 MobilenetV2；声学图像；光学图像；图像融合；水下目标分类
中图法分类号: TB566 文献标识码: A 文章编号: 1000-310X(2022)03-0462-09
DOI: 10.11684/j.issn.1000-310X.2022.03.017

Acoustic-optical image fusion underwater target classification method
based on improved MobilenetV2

GONG Wenjing 1,2,3 TIAN Jie 1,3 LI Baoqi 1,3 LIU Jiyuan 1,3

(1 Institute of Acoustics, Chinese Academy of Sciences, Beijing 100190, China)
(2 University of Chinese Academy of Sciences, Beijing 100049, China)
(3 Key Laboratory of Science and Technology on Advanced Underwater Acoustic Signal Processing, Chinese Academy of
Sciences, Beijing 100190, China)

Abstract: To solve the problems of low accuracy and high consumption of underwater target classification
under the condition of small samples, an acoustic-optical image fusion classification method is proposed. Firstly,
the Mobilenetv2 network is improved by removing the network layer after layer 9 and changing the channels
to 128. After the dimension reduction by flatten layer, a dense layer is added to get classification results.
Secondly, a fusion network structure is designed, which inputs acoustic image and optical image in pairs for
feature extraction. Then, the extracted feature maps are combined in the middle layer and used for target
classification. Using the real data to train the network, the results show that the improved Mobilenetv2 network
has better classification performance for underwater targets, and the accuracy of fusion network is improved
compared with that before fusion, indicating the applicability in underwater target classification.
Keywords: Improved Mobilenetv2; Acoustic image; Optical image; Image fusion; Underwater target
classification

2021-04-28 收稿; 2021-07-12 定稿
中国科学院国防科技重点实验室基金项目 (CXJJ-20S035)
∗
作者简介: 巩文静 (1996– ), 女, 山东泰安人, 博士研究生, 研究方向: 信号与信息处理。
† 通信作者 E-mail: ljy@mail.ioa.ac.cn