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

doi:10.7527/S1000-6893.2025.32356

Abstract

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

CLC Number:

V231.2

Baoquan SHI, Jiaming ZHOU, Wendong ZHANG, Xianmin CHEN, Qian HE. A network for fatigue crack segmentation and quantification using frequency-domain enhancement and linear attention mechanism[J]. Acta Aeronautica et Astronautica Sinica, 2025, 46(21): 532356.

Figures/Tables 20

Fig.1

Fig.2

Fig.3

Fig.4

Fig.5

Fig.6

Fig.7

Fig.8

Fig.9

Fig.10

Fig.11

Fig.12

Table 1

Table 2

Fig.13

Fig.14

Table 3

Table 4

Table 5

Table 6

References 34

[1]	WANG F. Application of deep learning algorithm in crack detection of green building materials［C］∥2022 6th Asian Conference on Artificial Intelligence Technology （ACAIT）. Piscataway： IEEE Press， 2022： 1-6.
[2]	WANG Y T， SHEN W W. Mechanical parts detection algorithm based on enhanced faster R-CNN［C］∥2021 China Automation Congress （CAC）. Piscataway： IEEE Press， 2021： 4348-4353.
[3]	DING K Y， DING Z H， ZHANG Z B， et al. SCD-YOLO： A novel object detection method for efficient road crack detection［J］. Multimedia Systems， 2024， 30（6）： 351.
[4]	LIU Z， ZHOU B. Research and application on the improved SSD chip defect inspection algorithm［C］∥2021 IEEE International Conference on Consumer Electronics and Computer Engineering （ICCECE）. Piscataway： IEEE Press， 2021： 551-555.
[5]	TAO H Q， LIU B X， CUI J Q， et al. A convolutional-transformer network for crack segmentation with boundary awareness［C］∥2023 IEEE International Conference on Image Processing （ICIP）. Piscataway： IEEE Press， 2023： 86-90.
[6]	ZHENG W W， JIANG X Y， FANG Z J， et al. TV-Net： A structure-level feature fusion network based on tensor voting for road crack segmentation［J］. IEEE Transactions on Intelligent Transportation Systems， 2024， 25（6）： 5743-5754.
[7]	ZHANG Q， CHEN S S， WU Y， et al. Improved U-Net network asphalt pavement crack detection method［J］. PLoS One， 2024， 19（5）： e0300679.
[8]	ZHOU H P， DENG B， SUN K L， et al. UTE-CrackNet： Transformer-guided and edge feature extraction U-shaped road crack image segmentation［J］. The Visual Computer， 2025， 41（4）： 2271-2283.
[9]	LIU H J， YANG J， MIAO X Y， et al. CrackFormer network for pavement crack segmentation［J］. IEEE Transactions on Intelligent Transportation Systems， 2023， 24（9）： 9240-9252.
[10]	SHAN J H， HUANG Y， JIANG W. DCUFormer： Enhancing pavement crack segmentation in complex environments with dual-cross/upsampling attention［J］. Expert Systems with Applications， 2025， 264： 125891.
[11]	LIU Z， LIN Y T， CAO Y， et al. Swin Transformer： Hierarchical vision transformer using shifted windows［C］∥2021 IEEE/CVF International Conference on Computer Vision （ICCV）. Piscataway： IEEE Press， 2021： 9992-10002.
[12]	HE K M， ZHANG X Y， REN S Q， et al. Deep residual learning for image recognition［C］∥2016 IEEE Conference on Computer Vision and Pattern Recognition （CVPR）. Piscataway： IEEE Press， 2016： 770-778.
[13]	PATRO B N， NAMBOODIRI V P， AGNEESWARAN V S. SpectFormer： Frequency and attention is what you need in a vision transformer［C］∥2025 IEEE/CVF Winter Conference on Applications of Computer Vision （WACV） . Piscataway： IEEE Press， 2017： 1800-1807.
[14]	SHI W Z， CABALLERO J， HUSZÁR F， et al. Real-time single image and video super-resolution using an efficient sub-pixel convolutional neural network［C］∥2016 IEEE Conference on Computer Vision and Pattern Recognition （CVPR）. Piscataway： IEEE Press， 2016： 1874-1883.
[15]	CHOLLET F. Xception： Deep learning with depthwise separable convolutions［C］∥2017 IEEE Conference on Computer Vision and Pattern Recognition （CVPR）. Piscataway： IEEE Press， 2017： 1800-1807.
[16]	GU A， DAO T. Mamba： Linear-time sequence modeling with selective state spaces［DB/OL］. arXiv preprint：2312. 00752， 2024.
[17]	HAN D C， WANG Z Y， XIA Z F， et al. Demystify Mamba in vision： A linear attention perspective［C］∥38th Conference on Neural Information Processing Systems （NeurIPS 2024）. Vancouver： OpenReview， 2024.
[18]	BLUM H. Biological shape and visual science （part I）［J］. Journal of Theoretical Biology， 1973， 38（2）： 205-287.
[19]	BAI X， LATECKI L J. Discrete skeleton evolution［M］∥Energy Minimization Methods in Computer Vision and Pattern Recognition. Berlin， Heidelberg： Springer， 2007： 362-374.
[20]	MONTANARI U. A method for obtaining skeletons using a Quasi-Euclidean distance［J］. Journal of the ACM， 1968， 15（4）： 600-624.
[21]	SHI Y， CUI L M， QI Z Q， et al. Automatic road crack detection using random structured forests［J］. IEEE Transactions on Intelligent Transportation Systems， 2016， 17（12）： 3434-3445.
[22]	YANG F， ZHANG L， YU S J， et al. Feature pyramid and hierarchical boosting network for pavement crack detection［J］. IEEE Transactions on Intelligent Transportation Systems， 2020， 21（4）： 1525-1535.
[23]	SHI Y， CUI L M， QI Z Q， et al. Automatic road crack detection using random structured forests［J］. IEEE Transactions on Intelligent Transportation Systems， 2016， 17（12）： 3434-3445.
[24]	ZOU Q， CAO Y， LI Q Q， et al. CrackTree： Automatic crack detection from pavement images［J］. Pattern Recognition Letters， 2012， 33（3）： 227-238.
[25]	ZOU Q， ZHANG Z， LI Q Q， et al. DeepCrack：Learning hierarchical convolutional features for crack detection［J］. IEEE Transactions on Image Processing， 2018， 28（3）： 1498-1512.
[26]	CHEN L C， ZHU Y K， PAPANDREOU G， et al. Encoder-decoder with atrous separable convolution for semantic image segmentation［C］∥Computer Vision-ECCV 2018. Cham： Springer， 2018： 833-851.
[27]	LIU Y H， YAO J， LU X H， et al. DeepCrack： A deep hierarchical feature learning architecture for crack segmentation［J］. Neurocomputing， 2019， 338： 139-153.
[28]	LAU S L H， CHONG E K P， YANG X， et al. Automated pavement crack segmentation using U-net-based convolutional neural network［J］. IEEE Access， 2020， 8： 114892-114899.
[29]	CHEN J， LU Y， YU Q， et al. TransUNet： Transformers make strong encoders for medical image segmentation［DB/OL］. arXiv preprint： 2102. 04306， 2021.
[30]	PANG J， ZHANG H， ZHAO H， et al. DcsNet： A real-time deep network for crack segmentation［J］. Signal， Image and Video Processing， 2022， 16（4）： 911-919.
[31]	WU H S， CHEN S H， CHEN G L， et al. FAT-Net： Feature adaptive transformers for automated skin lesion segmentation［J］. Medical Image Analysis， 2022， 76： 102327.
[32]	XIANG C， GUO J J， CAO R， et al. A crack-segmentation algorithm fusing transformers and convolutional neural networks for complex detection scenarios［J］. Automation in Construction， 2023， 152： 104894.
[33]	TAO H Q， LIU B X， CUI J Q， et al. A convolutional-transformer network for crack segmentation with boundary awareness［C］∥2023 IEEE International Conference on Image Processing （ICIP）. Piscataway： IEEE Press， 2023： 86-90.
[34]	ZHANG J M， ZENG Z G， SHARMA P K， et al. A dual encoder crack segmentation network with Haar wavelet-based high–low frequency attention［J］. Expert Systems with Applications， 2024， 256： 124950.

网络模型	Pr/%	Re/%	F1/%	mIoU/%
DeepLabv3+^［26］	72.82	74.58	73.69	61.33
DeepCrack^［27］	77.14	73.16	75.10	61.75
UNet-ResNet34^［28］	77.28	75.21	76.23	60.02
Swin Transformer^［11］	69.99	74.98	72.40	59.57
TransUNet^［29］	71.06	77.54	74.16	61.53
DscNet^［30］	79.32	74.47	76.82	63.11
FAT-Net^［31］	79.36	72.81	75.94	64.24
DTrc-Net^［32］	80.66	76.90	78.74	67.81
CT-CrackSeg^［33］	75.35	77.42	76.37	62.02
DECS-Net^［34］	79.84	78.61	79.22	62.83
CrackDAM-Net	83.11	77.56	80.24	68.05

网络模型	MAC/G	参数量/M	FPS
DeepLabv3+^［26］	22.14	59.23	30.64
DeepCrack^［27］	136.80	30.91	27.35
UNet-ResNet34^［28］	90.13	69.30	41.07
Swin Transformer^［11］	114.25	59.83	21.74
TransUNet^［29］	33.4	105.32	20.04
DscNet^［30］	15.75	25.85	39.19
FAT-Net^［31］	42.8	29.62	28.00
DTrc-Net^［32］	123.2	63.45	28.78
CT-CrackSeg^［33］	41.62	22.88	15.53
DECS-Net^［34］	26.08	68.86	30.02
CrackDAM-Net	25.24	63.89	37.36

编码器	Pr/%	Re/%	F1/%	mIoU/%
局部特征编码器	84.25	72.05	77.67	67.99
全局信息编码器	80.81	75.46	78.04	66.95
局部特征编码器+全局信息编码器	83.11	77.56	80.24	68.05

FDB加入的层数	Pr/%	Re/%	F1/%	mIoU/%
无FDB	80.53	75.13	77.74	66.97
仅第1层	83.39	76.83	79.98	67.39
1、2层	83.11	77.56	80.24	68.05
1、2、3层	82.35	76.34	79.23	67.19
1、2、3、4层	74.83	74.87	74.85	66.58

局部特征编码器	Pr/%	Re/%	F1/%	mIoU/%
ResNet-18	76.11	76.25	76.18	67.85
ResNet-34	80.42	76.44	78.38	67.91
ResNet-50	83.11	77.56	80.24	68.05
ResNet-101	83.61	75.18	79.17	67.65
ResNet-152	78.05	76.71	77.37	67.66

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

RichHTML

PDF (PC)

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 20

References 34

Related Articles 15

Recommended Articles

Metrics

Comments

序号	真实长度/mm	量化长度/mm	绝对误差/mm	相对误差/%
1	31.19	31.32	0.13	0.42
2	23.10	22.38	-0.72	-3.12
3	36.30	36.44	0.14	0.39
4	8.81	8.49	-0.32	-3.63
5	8.85	8.70	-0.15	-1.70
6	13.96	14.08	0.13	0.91
7	15.40	15.52	0.11	0.73
8	11.13	11.27	0.14	1.25
9	13.15	13.25	0.10	0.77
10	14.13	13.95	-0.18	-1.27

[1]	Lingjie MENG, Hongguang LI, Xinjun LI. SAR image simulation method guided by geomorphic category information [J]. Acta Aeronautica et Astronautica Sinica, 2025, 46(7): 331003-331003.
[2]	Guixian QU, Dongyang LIU, Xu YANG, Tian QIU, Chuankai LIU, Shuiting DING, Shuzheng YUAN, Kan GUO. Remaining useful life prediction method based on temporal information enhancement of sensors [J]. Acta Aeronautica et Astronautica Sinica, 2025, 46(17): 231634-231634.
[3]	Chuanyun WANG, Yang SU, Linlin WANG, Tian WANG, Jingjing WANG, Qian GAO. Multi-object continuous robust tracking algorithm for anti-UAV swarm [J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(7): 329017-329017.
[4]	Yueteng WU, Dun BA, Juan DU, Yunfei LI, Juntao CHANG. Compressor flow field reconstruction method based on deep attention networks [J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(24): 630580-630580.
[5]	Xiangxi KONG, Wenyuan QIN, Piaoyi SU, Yongzhao HUA, Xiwang DONG, Li WANG, Ying SU, Kun LYU. Damage assessment algorithm based on deep learning and fuzzy analytic hierarchy process [J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(19): 329503-329503.
[6]	Yonggang CHEN, Kangni LIU, Shuai WANG. Fault knowledge graph construction for aviation equipment based on BiGRU⁃Attention improvement [J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(18): 229916-229916.
[7]	Yuanhao HU, Yibo SONG, Jiahui LIU, Xiaoli ZHAO, Wenxiang DENG, Jian HU, Jianyong YAO. Fault diagnosis of electro-hydraulic proportional servo valves based on attention convolutional capsule networks [J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(15): 630407-630407.
[8]	Aoze YU, Weiwei WEI, Ping WANG, Jinqiang ZHANG, Wenxiong KE. Small target detection algorithm for UAV based on patch⁃wise co⁃attention [J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(14): 629148-629148.
[9]	Yanfang LIU, Jiayu SHE, Qiufan YUAN, Rui ZHOU, Naiming QI. Real⁃time small target detection networks for UAV remote sensing [J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(14): 630119-630119.
[10]	Junyu LI, Qiankun LIU, Ying FU. Infrared small object detection based on attention mechanism [J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(14): 628959-628959.
[11]	Jiqiang GAN, Xiaoping WANG. Surface defect detection of fiber placement based on virtual sample generation [J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(1): 428624-428624.
[12]	Zhiqiang FENG, Zhijun XIE, Zhengwei BAO, Kewei CHEN. Real⁃time dense small object detection algorithm for UAV based on improved YOLOv5 [J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2023, 44(7): 327106-327106.
[13]	Baohui JIA, Fan JIANG, Yuxin WANG, Du WANG. Fault diagnosis method based on civil aircraft maintenance text data [J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2023, 44(5): 326598-326598.
[14]	Guotao MAO, Tianmin DENG, Nanjing YU. Object detection in UAV images based on multi-scale split attention [J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2023, 44(5): 326738-326738.
[15]	Yi ZHANG, Yan ZHANG, Yu ZHANG, Yong ZHANG, Di LIU. Infrared aircraft target detection method based on multi-level feature enhancement fusion [J]. Acta Aeronautica et Astronautica Sinica, 2023, 44(22): 629220-629220.