基于并行注意力UNet的裂缝检测方法

刘凡; 王君锋; 陈峙宇; 许峰

doi:10.7544/issn1000-1239.2021.20210335

基于并行注意力UNet的裂缝检测方法

(河海大学计算机信息学院南京 210098) (海岸灾害与防护教育部重点实验室(河海大学) 南京 210098) (191307040025@hhu.edu.cn)

基金项目: 江苏省自然科学基金项目(BK20191298)；中央高校基本科研业务费专项资金(B200202175)；河海大学海岸灾害及防护教育部重点实验室开放基金项目(20150009)

详细信息

中图分类号: TP391
计量
- 文章访问数: 766
- HTML全文浏览量: 19
- PDF下载量: 560
出版历程
- 发布日期: 2021-07-31

Parallel Attention Based UNet for Crack Detection

(College of Computer Information, Hohai University, Nanjing 210098) (Key Laboratory of Coastal Disaster and Protection(Hohai University), Ministry of Education, Nanjing 210098)

Funds: This work was supported by the Natural Science Foundation of Jiangsu Province (BK20191298), the Fundamental Research Funds for the Central Universities (B200202175), and the Open Project of the Key Laboratory of Coastal Disaster and Protection (Hohai University) of Ministry of Euducation (20150009).

摘要

摘要: 裂缝对公共设施而言存在着安全隐患，因此裂缝检测是公共设施进行维护的重要手段.由于裂缝图像中存在噪声、光线、阴影等因素干扰，神经网络在训练时极易被影响，导致预测结果出现偏差，降低预测效果.为减少这些干扰，设计了一个并行注意力机制，并将其嵌入到UNet网络的解码部分，进而提出了并行注意力UNet(parallel attention based UNet, PA-UNet).该方法分别从通道和空间2个维度加大裂缝特征权重以抑制干扰，然后对这2个维度生成的特征进行融合，以获得更具互补性的裂缝特征.为了验证该方法的有效性，选取了4个数据集进行实验，结果表明该方法较现有的主流方法，裂缝检测效果更加优异.同时，为了验证并行注意力机制的有效性，选取了4种注意力机制与其进行对比实验，结果表明并行注意力机制效果优于其他注意力机制.
- 裂缝检测 /
- 并行注意力机制 /
- UNet /
- 抑制干扰 /
- 互补性
Abstract: Cracks have hidden safety hazards to public facilities, so crack detection is essential for the maintenance of public facilities. Due to the interference of noise, light, shadow, and other factors in the crack images, the neural network is easily affected during the training process, which causes deviations in the prediction results and reduces the prediction effect. To suppress these disturbances, a parallel attention mechanism is designed and then the parallel attention based UNet(PA-UNet) is proposed by embedding this attention mechanism into UNet. The parallel attention mechanism increases the weights of crack features from the two dimensions of channel and space to suppress interference, then fuses the features generated by these two dimensions to obtain more complementary crack features. To verify the effectiveness of the proposed method, we have conducted experiments on four data sets. Experimental results show that our method outperforms the existing popular methods. Meanwhile, to demonstrate the effectiveness of the parallel attention mechanism, we conduct a comparative experiment with other four attention mechanisms. The results show that the parallel attention mechanism performs better than others.
- crack detection /
- parallel attention mechanism /
- UNet /
- suppress interference /
- complementary

HTML全文

参考文献(0)

施引文献(14)

期刊类型引用(7)

1.	魏波，冯乃勤. 基于入侵诱骗的网络拓扑污染攻击防御研究. 计算机仿真. 2024(05): 410-414 . 百度学术
2.	金柯君，于洪涛，吴翼腾，李邵梅，张建朋，郑洪浩. 改进的基于奇异值分解的图卷积网络防御方法. 计算机应用. 2023(05): 1511-1517 . 百度学术
3.	宋国顺. 基于特征加权聚合的传感网络多模式攻击检测方法. 通化师范学院学报. 2023(10): 74-80 . 百度学术
4.	金柯君，于洪涛，李邵梅，张建朋. 基于注意力机制的图卷积网络防御方法. 信息工程大学学报. 2023(06): 718-724 . 百度学术
5.	刘勇. 基于图算法的定向越野数据智能分析系统设计. 自动化与仪器仪表. 2022(08): 159-164 . 百度学术
6.	张瑾，朱桂祥，王宇琛，郑烁佳，陈镜潞. 基于异质图表达学习的跨境电商推荐模型. 电子与信息学报. 2022(11): 4008-4017 . 百度学术
7.	吴翼腾，刘伟，于洪涛. 图神经网络的标签翻转对抗攻击. 通信学报. 2021(09): 65-74 . 百度学术