Spacecraft Recovery & Remote Sensing, Volume. 46, Issue 2, 146(2025)
A Bathymetric Algorithm of Laser Echo Based on CNN-LSTM
Figures & Tables(10)
![[in Chinese]](/Images/icon/loading.gif){kind=icon}
Table 1. Initial setup of CNN-LSTM model
View tableView in ArticleTable 1. Initial setup of CNN-LSTM model
参数 名称 批量 大小 学习率 优化器 训练 轮数 学习率 衰减率 学习率衰减的 步长 初始 动量值 动量 衰减率 动量衰减 的步长 正则化项 参数 指标值 16 0.01 Adam 400 0.5 20 0.1 0.5 20 0.5
Table 2. Experimental results in testing datasets
View tableView in ArticleTable 2. Experimental results in testing datasets
模型名称 最佳轮数 准确率/% 损失 水面点预测评估 水底点预测评估 P/% R/% F1/% P/% R/% F1/% CNN 4 96.11 0.136 8 56.08 74.96 64.16 79.19 93.67 85.82 LSTM 58 94.99 0.103 5 52.86 26.29 35.12 74.06 64.82 69.13 1D FCN 41 95.89 0.106 6 53.73 30.40 42.10 92.92 85.65 87.08 CNN-LSTM 216 97.70 0.085 9 71.81 85.47 78.05 85.92 95.02 90.24
Table 3. Experimental results in validation datasets
View tableView in ArticleTable 3. Experimental results in validation datasets
模型名称 正确率/% 损失 水面点预测评估 水底点预测评估 RMSE/m P/% R/% F1/% P/% R/% F1/% CNN 96.02 0.1374 55.32 74.09 63.34 78.92 93.31 85.51 0.72 LSTM 94.93 0.1048 52.27 26.23 34.93 72.70 64.58 68.40 1.37 1D FCN 95.89 0.1062 54.62 30.64 39.26 92.33 84.54 88.26 1.26 CNN-LSTM 97.62 0.0855 70.72 85.22 77.30 85.93 94.33 89.93 0.46
Tools
Get Citation
Copy Citation Text
Li SHENG, Peize LI, Yangrui XU, Junnan BIAN, Kun LIANG. A Bathymetric Algorithm of Laser Echo Based on CNN-LSTM[J]. Spacecraft Recovery & Remote Sensing, 2025, 46(2): 146
Paper Information
Category:
Received: Oct. 16, 2024
Accepted: Oct. 21, 2024
Published Online: May. 23, 2025
The Author Email: Kun LIANG (liangkun@hust.edu.cn)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20