双重细化门控自适应融合的道路裂缝检测算法

doi:10.3778/j.issn.1673-9418.2501001

摘要/Abstract

摘要： 道路裂缝存在缺陷类型多样、异常区域复杂等特点，当前的目标检测算法在通道和空间维度存在冗余的特征处理、阶段信息的盲目式融合等问题，导致网络效率低、关键信息丢失。提出了一种双重细化门控自适应融合网络（DR-DETR），在潜在空间中实现特征在通道和空间维度上的双重精细化处理，解决上述问题的同时，也提高了道路裂缝的检测精度。构建一种通道-空间双重细化的信息蒸馏机制，分别对通道和空间维度内冗余的特征进行信息蒸馏，减少网络对特征的冗余处理，实现关键性特征的高效表征；针对阶段性特征的粗粒度融合，提出了特征信息门控自适应融合模块（FGAF-Fusion），利用增强型条带大核卷积获取全局信息，借助对比感知注意力实现通道间的交互和融合，同时利用门控自适应融合机制筛选出关键性的小目标语义信息；设计Res-DCNv3模块，利用DCNv3可变形卷积的灵活性来精确提取形态各异的道路裂缝特征。在RDD2022公开数据集中的实验结果显示，提出的DR-DETR在mAP0.50和mAP0.50:0.95分别达到了51.7%和24.9%，相较于RT-DETR分别提升了4.2和3.3个百分点。在道路裂缝目标检测任务中，提出的DR-DETR可以有效检测出不同类型的道路缺陷，展现出极具竞争性的检测结果和良好的鲁棒性。

关键词: 缺陷检测, 双重细化, 门控自适应融合, RT-DETR, 可变形卷积

Abstract: The road crack detection task presents various defect types and complex anomaly regions. The current object detection algorithms suffer from redundant feature processing in channel and spatial dimensions, blind fusion of stage information, and other issues, leading to low network efficiency and loss of crucial information. This paper introduces a dual-refinement gate-controlled adaptive fusion network (DR-DETR) that achieves dual-refinement processing of features in channel and spatial dimensions in latent space, addressing previous issues and enhancing road crack detection accuracy. This paper constructs a dual-refinement information distillation mechanism in channel and spatial dimensions, distills redundant features separately in channel and spatial dimensions, reduces redundant processing of features by the network, and achieves efficient representation of crucial features. Aiming at the coarse-grained fusion of stage features, a feature-gated fine-grained adaptive fusion module (FGAF-Fusion) is proposed, which utilizes enhanced stripe-wise dilated convolutions to capture global information, then employs contrastive perceptual attention for inter-channel interaction and fusion, while utilizing a gate-adaptive fusion mechanism to select critical semantic information of small targets. This paper designs the Res-DCNv3 module, utilizing the flexibility of DCNv3 deformable convolutions to accurately extract diverse features of road cracks with varying morphologies. Experimental results on the RDD2022 public dataset demonstrate that the proposed DR-DETR achieves mAP0.50 and mAP0.50:0.95 of 51.7% and 24.9%, respectively, representing improvements of 4.2 and 3.3 percentage points over RT-DETR. In road crack object detection tasks, the proposed DR-DETR can effectively detect various types of road defects, demonstrating highly competitive detection results and good robustness.

Key words: defect detection, dual refinement, gate-controlled adaptive fusion, RT-DETR, deformable , convolution

冯永安, 张紫扬, 张旭. 双重细化门控自适应融合的道路裂缝检测算法[J]. 计算机科学与探索, 2025, 19(11): 2981-2993.

FENG Yong'an, ZHANG Ziyang, ZHANG Xu. Dual-Refinement Gate-Controlled Adaptive Fusion Algorithm for Road Crack Detection[J]. Journal of Frontiers of Computer Science and Technology, 2025, 19(11): 2981-2993.

参考文献

[1] 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.
[2] FAN L L, WANG D D, WANG J H, et al. Pavement defect detection with deep learning: a comprehensive survey[J]. IEEE Transactions on Intelligent Vehicles, 2024, 9(3): 4292-4311.
[3] GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2014: 580-587.
[4] CAI Z W, VASCONCELOS N. Cascade R-CNN: high quality object detection and instance segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 43(5): 1483-1498.
[5] LIANG F T, ZHOU Y, CHEN X, et al. Review of target detection technology based on deep learning[C]//Proceedings of the 5th International Conference on Control Engineering and Artificial Intelligence. New York: ACM, 2021: 132-135.
[6] REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[C]//Advances in Neural Information Processing Systems 28, 2015: 91-99.
[7] ZHANG K H, SHEN H K. Solder joint defect detection in the connectors using improved Faster-RCNN algorithm[J]. Applied Sciences, 2021, 11(2): 576.
[8] YANG A M, JIANG T Y, HAN Y, et al. Research on application of on-line melting in-SITU visual inspection of iron ore powder based on Faster R-CNN[J]. Alexandria Engineering Journal, 2022, 61(11): 8963-8971.
[9] HE K M, GKIOXARI G, DOLLáR P, et al. Mask R-CNN[C]//Proceedings of the 2017 IEEE International Conference on Computer Vision. Piscataway: IEEE, 2017: 2980-2988.
[10] JIANG P Y, ERGU D, LIU F Y, et al. A review of Yolo algorithm developments[J]. Procedia Computer Science, 2022, 199: 1066-1073.
[11] REDMON J, FARHADI A. YOLOv3: an incremental improvement[EB/OL]. [2024-11-20]. https://arxiv.org/abs/1804.02767.
[12] WANG C Y, YEH I H, MARK LIAO H Y. YOLOv9: learning what you want to learn using programmable gradient information[C]//Proceedings of the 18th European Conference on Computer Vision. Cham: Springer, 2024: 1-21.
[13] 任晶晶, 徐志远. 基于YOLOv8的轻量化道路裂缝检测模型[J]. 山西电子技术, 2024(4): 54-56.
REN J J, XU Z Y. A lightweight road crack detection model based on YOLOv8[J]. Shanxi Electronic Technology, 2024(4): 54-56.
[14] 蒋大伟, 吴正平, 景思伟. 基于改进YOLOv5的道路缺陷检测与分类研究[J]. 信息技术与信息化, 2024(2): 31-34.
JIANG D W, WU Z P, JING S W. Research on road defect detection and classification based on improved YOLOv5[J]. Information Technology and Informatization, 2024(2): 31-34.
[15] 杨豪, 刘李彦, 张军辉, 等. 环境适应性优化的轻量化多尺度道路裂缝检测[J]. 中国公路学报, 2025, 38(7): 118-134.
YANG H, LIU L Y, ZHANG J H, et al. Environmental adaptability optimization lightweight multi-scale road crack detection[J]. China Journal of Highway and Transport, 2025, 38(7): 118-134.
[16] 宣以国, 余成波, 蒋启超, 等. 基于改进YOLOv7的道路裂缝和坑洞检测算法[J]. 科学技术与工程, 2024, 24(17): 7205-7213.
XUAN Y G, YU C B, JIANG Q C, et al. Improved YOLOv7 road crack and pothole detection algorithm[J]. Science Technology and Engineering, 2024, 24(17): 7205-7213.
[17] 侯涛, 张田明, 牛宏侠. MFF-YOLO: 多尺度特征融合的轻量级道路缺陷检测算法[J/OL]. 工程科学与技术 [2024-11-20]. https://kns.cnki.net/kcms/detail/51.1773.TB.20241115. 1437.003.html.
HOU T, ZHANG T M, NIU H X. MFF-YOLO: a lightweight road damage detection algorithm based on multiscale feature fusion[J/OL]. Advanced Engineering Sciences[2024-11-20]. https://kns.cnki.net/kcms/detail/51.1773.TB. 20241115.1437.003.html.
[18] 李胜杰, 刘贵如, 王陆林, 等. 一种轻量化YOLOv5道路缺陷检测算法[J/OL]. 天津理工大学学报 [2024-11-20]. https:// kns.cnki.net/kcms/detail/12.1374.N.20241101. 1754.006.html.
LI S J, LIU G R, WANG L L, et al. A lightweight YOLOv5 road defect detection algorithm[J/OL]. Journal of Tianjin University of Technology [2024-11-20]. https://kns.cnki.net/kcms/detail/12.1374.N.20241101.1754.006.html.
[19] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Advances in Neural Information Processing Systems 30, 2017: 5998-6008.
[20] CARION N, MASSA F, SYNNAEVE G, et al. End-to-end object detection with transformers[C]//Proceedings of the 16th European Conference on Computer Vision. Cham: Springer, 2020: 213-229.
[21] MENG D P, CHEN X K, FAN Z J, et al. Conditional DETR for fast training convergence[C]//Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision. Piscataway: IEEE, 2021: 3631-3640.
[22] ZHU X Z, SU W J, LU L W, et al. Deformable-DETR: deformable transformers for end-to-end object detection[EB/OL]. [2024-11-20]. https://arxiv.org/abs/2010.04159.
[23] LI F, ZHANG H, LIU S L, et al. DN-DETR: accelerate DETR training by introducing query denoising[C]//Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2022: 13609-13617.
[24] LIU S L, LI F, ZHANG H, et al. DAB-DETR: dynamic anchor boxes are better queries for DETR[EB/OL]. [2024-11-21]. https://arxiv.org/abs/2201.12329.
[25] 许正森, 雷相达, 管海燕. 多尺度局部特征增强Transformer道路裂缝检测模型[J]. 中国图象图形学报, 2023, 28(4): 1019-1028.
XU Z S, LEI X D, GUAN H Y. Multi-scale local feature enhanced transformer network for pavement crack detection[J]. Journal of Image and Graphics, 2023, 28(4): 1019-1028.
[26] 刘知阳. 基于Transformer的道路裂缝检测研究[D]. 汉中: 陕西理工大学, 2024.
LIU Z Y. Research on road crack detection based on transformer[D]. Hanzhong: Shaanxi University of Technology, 2024.
[27] ZHAO Y A, LV W Y, XU S L, et al. DETRs beat YOLOs on real-time object detection[C]//Proceedings of the 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2024: 16965-16974.
[28] ZHANG H, WU C R, ZHANG Z Y, et al. ResNeSt: split-attention networks[C]//Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2022: 2735-2745.
[29] HU J, SHEN L, SUN G. Squeeze-and-excitation networks[EB/OL]. [2024-11-21]. https://arxiv.org/abs/1709.01507.
[30] CHEN J R, KAO S H, HE H, et al. Run, don??t walk: chasing higher FLOPS for faster neural networks[C]//Proceedings of the 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2023: 12021-12031.
[31] CHEN Y P, FAN H Q, XU B, et al. Drop an octave: reducing spatial redundancy in convolutional neural networks with octave convolution[C]//Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision. Piscataway: IEEE, 2019: 3434-3443.
[32] ZHU L, WANG X J, KE Z H, et al. BiFormer: vision transformer with bi-level routing attention[C]//Proceedings of the 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2023: 10323-10333.
[33] LAU K W, PO L M, REHMAN Y A U. Large separable kernel attention: rethinking the large kernel attention design in CNN[J]. Expert Systems with Applications, 2024, 236: 121352.
[34] CHOLLET F. Xception: deep learning with depthwise separable convolutions[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2017: 1800-1807.
[35] ZHANG H, LI F, LIU S L, et al. DINO: DETR with improved denoising anchor boxes for end-to-end object detection[EB/ OL]. [2024-11-21]. https://arxiv.org/abs/2203.03605.
[36] ARYA D, MAEDA H, GHOSH S K, et al. RDD2022: a multi-national image dataset for automatic road damage detection[J]. Geoscience Data Journal, 2024, 11(4): 846-862.
[37] ZHANG H Y, WANG Y, DAYOUB F, et al. VarifocalNet: an IoU-aware dense object detector[C]//Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2021: 8514-8523.