航空学报 > 2025, Vol. 46 Issue (21): 532356-532356   doi: 10.7527/S1000-6893.2025.32356

中国飞机强度研究所建所 60 周年专刊

融合频域增强与线性注意力机制的疲劳裂纹分割与量化网络

史宝全1,2,3(), 周嘉明1, 张文东3,4, 陈先民3,4, 贺谦3,4   

  1. 1.西安电子科技大学 机电工程学院,西安 710071
    2.西安宝创速维智能科技有限公司,西安 712000
    3.强度与结构完整性全国重点实验室,西安 710065
    4.中国飞机强度研究所,西安 710065
  • 收稿日期:2025-06-03 修回日期:2025-06-17 接受日期:2025-07-17 出版日期:2025-07-28 发布日期:2025-07-25
  • 通讯作者: 史宝全 E-mail:bqshi@xidian.edu.cn
  • 基金资助:
    国家重点研发计划(2023YFF0720400);强度与结构完整性全国重点实验室开放基金(LSSIKFJJ202402005);陕西省技术创新引导专项(2024ZC-YYDP-28)

A network for fatigue crack segmentation and quantification using frequency-domain enhancement and linear attention mechanism

Baoquan SHI1,2,3(), Jiaming ZHOU1, Wendong ZHANG3,4, Xianmin CHEN3,4, Qian HE3,4   

  1. 1.School of Mechano-Electronic Engineering,Xidian University,Xi’an,710071,China
    2.Xi’an Baochuang Suwei Intelligent Research Co. ,Ltd,Xi’an,712000,China
    3.National Key Laboratory of Strength and Structural Integrity,Xi’an,710065,China
    4.Aircraft Strength Research Institute of China,Xi’an,710065,China
  • Received:2025-06-03 Revised:2025-06-17 Accepted:2025-07-17 Online:2025-07-28 Published:2025-07-25
  • Contact: Baoquan SHI E-mail:bqshi@xidian.edu.cn
  • Supported by:
    National Key R&D Program of China (MOST)(2023YFF0720400);The Open Fund of National Key Laboratory of Strength and Structural Integrity(LSSIKFJJ202402005);Shaanxi Province Technology Innovation Guidance Program(2024ZC-YYDP-28)

摘要:

疲劳裂纹是导致工程结构失效的主要诱因之一。针对结构强度试验中疲劳裂纹精准检测的实际需求,提出一种融合频域增强与Mamba线性注意力机制的疲劳裂纹分割与量化网络——CrackDAM-Net。该网络采用局部特征编码器和全局信息编码器双分支并行结构,同时兼顾了疲劳裂纹图像的局部细节特征与全局上下文信息。首先,设计了一种融合频域增强与Mamba线性注意力的全局信息编码器:前2层为混合频域-注意力层,后2层为纯注意力层。该编码器不仅能够有效提取裂纹图像的全局特征,还能显著降低计算复杂度。其次,提出了一种混合空间注意力、通道注意力以及像素注意力机制的多尺度特征融合模块,能够从多个维度充分挖掘和整合局部特征编码器与全局信息编码器之间的互补信息,提升特征表达能力。最后,建立了一种结合正交投影与离散骨架演化的裂纹长度量化方法,以实现结构强度试验中疲劳裂纹长度的精准测量。实验结果表明:CrackDAM-Net在精度与速度间取得良好平衡,其分割精确度、召回率和F1分数分别达到83.11%、77.56%和80.24%,分割速度达到37帧/s,疲劳裂纹长度量化误差在±4%以内,展现出较强的工程应用潜力。

关键词: 疲劳裂纹, 分割网络, 注意力机制, 特征融合, 自动量化

Abstract:

Fatigue cracks are one of the primary causes of structural failure in engineering structures. To meet the practical need for precise detection of fatigue cracks in structural strength testing, we propose a fatigue crack segmentation and quantification network-CrackDAM-Net-based on frequency domain enhancement and the Mamba linear attention mechanism. This network adopts a dual-branch parallel structure comprising a local feature encoder and a global information encoder, effectively balancing the local detailed features and global contextual information of fatigue crack images. A global information encoder combining frequency domain enhancement and Mamba linear attention is designed: the first two layers are hybrid frequency domain-attention layers, and the last two layers are pure attention layers. This encoder not only effectively extracts global features from crack images but also significantly reduces computational complexity. A multi-scale feature fusion module combining spatial attention, channel attention, and pixel attention mechanisms is proposed, which can fully mine and integrate complementary information between the local feature encoder and the global information encoder from multiple dimensions to enhance feature expression capabilities. A crack length automatic quantification method combining orthogonal projection and discrete skeleton evolution is established to achieve precise measurement of fatigue crack length in structural strength tests. Experimental results demonstrate that CrackDAM-Net achieves a good balance between accuracy and speed, with segmentation accuracy, recall rate, and F1 score reaching 83.11%, 77.56%, and 80.24%, respectively, segmentation speed reaching 37 frame/s, and fatigue crack length quantification error within ±4%, showcasing strong potential for engineering applications.

Key words: fatigue cracks, partitioning network, attention mechanism, feature fusion, automatic quantification

中图分类号: