Tunnel leakage detection method based on texture features

Table 1. Experimental environment and parameter configu⁃ration
View table
View in Article
Table 1. Experimental environment and parameter configu⁃ration
硬件/软件参数/版本
操作系统 Ubuntu 18.04.6
CPU Intel（R）Core（TM）
GPU NVIDIA GeForce RTX 3080 Lite Hash Rate
内存 31 GB
Anaconda 4.10.1
PyTorch 1.11.0
python 3.9

Table 2. Hyperparameters
View table
View in Article
Table 2. Hyperparameters
Batch size Momentum Lr Optimizer Epoch Loss
4 0.9 5e-4 Adam 100
CrossEntropy
loss

Table 3. Performance comparison of different models

View table

View in Article

Table 3. Performance comparison of different models

模型	Accuracy/%	F1-score/%	Precision/%	Recall/%	IoU/%	FLOPs/G	Params/M
U-Net	94.23	88.54	89.30	87.81	78.67	63.09	12.14
DeepLabv3+	93.13	86.95	87.56	86.36	77.82	694.61	27.54
BiseNetv2	93.98	87.79	88.42	87.17	78.25	17.53	3.25
Swin	94.57	88.82	89.48	88.17	79.92	185.45	31.73
SegNet	94.20	88.49	89.25	87.75	78.59	565.41	29.44
本文算法	94.68	89.30	89.91	88.71	80.64	826.73	29.76

Table 4. Performance comparison
View table
View in Article
Table 4. Performance comparison
LTC LCA Accuracy/% IoU/%
- - 89.92 76.68
√ - 93.46 79.21
- √ 92.19 78.93
√ √ 94.68 80.64

Tools

Get Citation

Copy Citation Text

Yue YANG, Yangyang CHEN, Xudong KANG, Bin WU. Tunnel leakage detection method based on texture features[J]. Chinese Journal of Liquid Crystals and Displays, 2025, 40(9): 1347

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites