As key load-bearing elements, bridge piers are highly sensitive to concrete construction quality, which directly affects structural safety and durability. Surface defects such as color differences and bugholes are common and, if not properly assessed, may accelerate deterioration. To overcome the limitations of conventional inspection methods in precision, efficiency, and objectivity, this study proposes an intelligent framework integrating high-resolution panoramic imaging, deep semantic segmentation, and spatial color-difference analysis. The main contributions include: (1) A segment-based matrix image acquisition strategy is developed using unmanned aerial vehicles, combined with a DSeam-driven deep stitching algorithm to generate panoramic images of piers with geometric consistency and luminance continuity, serving as a unified visual basis for defect recognition. (2) For bughole segmentation, a PoreMamba-Net semantic network based on a state space modeling architecture is constructed. The network employs multi-branch convolution and selective scanning mechanisms to enhance fine-grained perception and contextual representation of small targets, enabling high-precision segmentation. A quantitative evaluation system composed of the apparent bughole area ratio, maximum bughole diameter, and bughole distribution standard deviation is further proposed to comprehensively characterize bughole quantity, size, and spatial uniformity. (3) For unbounded and non-structural color difference defects, a pixel-level analysis method is proposed by combining multi-scale retinex with color restoration-based illumination normalization and the CIE94 color difference model. Moran’s I spatial autocorrelation index is introduced to jointly quantify color deviation intensity and spatial consistency, supporting classification of color differences and detection of clustered anomalies. The proposed framework is validated on typical piers from the Shiziyang Channel Bridge construction project in Guangdong Province. Results demonstrate strong adaptability, robustness, and accuracy in complex field conditions, providing reliable technical support for concrete quality acceptance and lifecycle management.
VikanH. State of the art – quality of concrete surfaces. SINTEF Building and Infrastructure, Report SBF BK A07013. Trondheim, Norway, 2007, p. 36.
2.
American Society of Concrete Contractors (ASCC). ASCC position statement #8: Bugholes in formed concrete. American Society of Concrete Contractors, St. Louis, MO, August2011.
3.
ACI Committee 347. ACI 347.3R—guide to formed concrete surfaces. American Concrete Institute, Farmington Hills, MI, 2013.
4.
HuLYoshitakeISayamaM, et al. Effect of surface bug-holes on chloride penetration of concrete coated with penetrants. Mag Concr Res2020; 74(5): 217–224.
5.
NRMCA. CIP 23 — discoloration. National Ready Mixed Concrete Association, Silver Spring (MD), USA. [Note: one-page sheet, access via NRMCA website].
6.
PanigatiTZiniMStriccoliD, et al. Drone-based bridge inspections: current practices and future directions. Autom Constr2025; 173: 106101.
7.
MaoJHeZHeY, et al. A review of effect of mineral admixtures on appearance quality of fair-faced concrete and techniques for their measurement. Sustainability2023; 15(19): 14623.
8.
ShyhaIRichardsonACoventryK, et al. Mould release technologies with regard to concrete surface finish. Mag Concr Res2015; 68(2): 87–98.
9.
ShiJQianCHaoZ, et al. Effects of equipment parameters, ambient parameters and concrete surface moisture on the acquisition of fair-faced concrete appearance quality information. J Build Eng2023; 72: 106476.
10.
JungHJLeeJHYoonS, et al. Bridge inspection and condition assessment using unmanned aerial vehicles (UAVs): major challenges and solutions from a practical perspective. Smart Struct Syst2019; 24(5): 669–681.
11.
ZhangJHChenWZhangJ. UAV-based quantitative crack measurement for bridges integrating four-point laser metric calibration and mamba segmentation. Autom Constr2026; 182: 106774.
12.
ChenWZhangXYuanB, et al. Adaptive drone inspection strategy for bridge based on multi-level representation learning. Adv Eng Informatics2025; 68: 103589.
13.
AvendañoJCLeanderJKaroumiR, et al. Image-based inspection of concrete cracks using UAV photography. In: Proceedings of the 11th international conference on bridge maintenance, safety and management (IABMAS 2022), Barcelona, Spain, 11–15 July 2022, pp. 1973–1978. London: CRC Press.
14.
XieRYaoJLiuK, et al. Automatic multi-image stitching for concrete bridge inspection by combining point and line features. Autom Constr2018; 90: 265–280.
15.
LiRGaoPCaiX, et al. A real-time incremental video mosaic framework for UAV remote sensing. Remote Sens2023; 15(8): 2127.
16.
ChenWZhangJ. Integrating dual-modal camera systems on unmanned aerial vehicles for bridge tower defect detection. J Civil Struct Health Monit2025; 15(4): 123–137.
17.
WeiWDingLLuoH, et al. Automated bughole detection and quality performance assessment of concrete using image processing and deep convolutional neural networks. Constr Build Mater2021; 281: 122576.
18.
MalischWRBrownHJ. Examination and evaluation of ACI 347.3R-13 ‘Guide to Formed Concrete Surfaces’. Final report to the ASCC Education, Research & Development Foundation, RMC Research & Education Foundation, and ACI Foundation, 2017, p. 145.
19.
American Society of Concrete Contractors. Bugholes in formed concrete. ASCC Position Statement #8. St. Louis, MO: American Society of Concrete Contractors, 2011, p. 1.
20.
LiuBYangT. Image analysis for detection of bugholes on concrete surface. Constr Build Mater2017; 137: 432–440.
21.
WuYHuangZZhangJ, et al. Grouting defect detection of bridge tendon ducts using impact echo and deep learning via a two-stage strategy. Mech Syst Signal Process2025; 235: 112955.
22.
WuYCaiDGuS, et al. Compressive strength prediction of sleeve grouting materials in prefabricated structures using hybrid optimized XGBoost models. Constr Build Mater2025; 476: 141319.
23.
SunYYangYYaoG, et al. Autonomous crack and bughole detection for concrete surface image based on deep learning. IEEE Access2021; 9: 85709–85720.
24.
WeiFYaoGYangY, et al. Instance-level recognition and quantification for concrete surface bughole based on deep learning. Autom Constr2019; 107: 102920.
25.
QianCDuW. Analysis method of apparent quality of fair-faced concrete based on convolutional neural network machine learning. J Build Eng2023; 80: 108154.
26.
WeiYWeiZXueK, et al. Automated detection and segmentation of concrete air voids using zero-angle light source and deep learning. Autom Constr2021; 130: 103877.
27.
XiaZWuBChanCY, et al. Deep-learning-based pyramid-transformer for localized porosity analysis of hot-press sintered ceramic paste. PLoS One2024; 19(9): e0306385.
28.
RaoASNguyenTLeST, et al. Attention recurrent residual U-Net for predicting pixel-level crack widths in concrete surfaces. Struct Health Monit2022; 21(6): 2732–2749.
29.
ZhaoWLiuYZhangJ, et al. Automatic pixel-level crack detection and evaluation of concrete structures using deep learning. Struct Control Health Monit2022; 29(8): e2981.
30.
ZhangWJWanHPToddMD. An efficient 2D–3D fusion method for bridge damage detection under complex backgrounds with imbalanced training data. Adv Eng Inform2025; 65: 103373.
31.
ChuHGaiJChenW, et al. CBRFormer: rendering technology-based transformer for refinement segmentation of bridge crack images. Adv Eng Inform2026; 69: 103868.
32.
ChuHChenWDengL. Refined segmentation of high-resolution bridge crack images via probability map-guided point rendering technique. Comput Aided Civ Infrastruct Eng2025; 40(27): 4946–4969.
33.
LiuHJiaCShiF, et al. SCSegamba: lightweight structure-aware vision mamba for crack segmentation in structures. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (CVPR), Nashville, TN, USA, 11–15 June 2025, pp. 29406–29416. New York, NY: IEEE.
34.
NikoueiMBaroutianBNabaviS, et al. Small object detection: a comprehensive survey on challenges, techniques and real-world applications. Intelli Syst Appl2025; 27: 200561.
35.
GuADaoT. Mamba: linear-time sequence modeling with selective state spaces. In: Proceedings of the 1st Conference on Language Modeling (COLM), Philadelphia, PA, July 15–17, 2024.
36.
ZhuLLiaoBZhangQ, et al. Vision Mamba: efficient visual representation learning with bidirectional state space model. In: Proceedings of the 41st international conference on machine learning (ICML), Vienna, Austria, July 21–27, 2024, pp. 62429–62442. Cambridge, MA: PMLR.
37.
TangWWuZWangW, et al. VM-UNet++ research on crack image segmentation based on improved VM-UNet. Sci Rep2025; 15: 8938.
38.
ZuoXShengYShenJ, et al. Topology-aware Mamba for crack segmentation in structures. Autom Constr2024; 168: 105845.
39.
HanCYangHYangY. Enhancing pixel-level crack segmentation with visual mamba and convolutional networks. Autom Constr2024; 168: 105770.
40.
YaoGWeiFYangY, et al. Deep-learning-based bughole detection for concrete surface image. Adv Civ Eng2019; 2019: 8582963.
41.
ACI Committee 347. Guide to formed concrete surfaces (ACI 347.3R-13). Farmington Hills, MI: American Concrete Institute, 2014.
42.
JiangQYuCZhouM, et al. Aesthetic properties of concrete surfaces: an experimental study employing three typical formworks with emphasis on reusability. Case Stud Constr Mater2025; 22: e04885.
43.
RoySKThyeLBNorthwoodDO. The evaluation of paint performance for exterior applications in Singapore’s tropical environment. Build Environ1996; 31(5): 477–486.
44.
JiangYPangDLiC. A deep learning approach for fast detection and classification of concrete damage. Autom Constr2021; 128: 103785.
45.
KárolyfiKAHorváthAPappF. Digital image processing method for evaluation of discoloration on fair-faced concrete surfaces. Építőanyag – J Silicate Compos Mater2022; 74(2): 67–70.
46.
LuJLengFXiaJ, et al. Quantitative evaluation and traceability analysis of concrete apparent shade and apparent uniformity. Case Stud Constr Mater2025; 23: e05006.
47.
LiuQLiaoZXiongY, et al. Effect of imitation fair-faced curing and protective coating on the durability of concrete. J Build Eng2023; 63: 105540.
48.
YaoGSunWYangY, et al. Chromatic aberration identification of fair-faced concrete research based on multi-scale lightweight structured data algorithm. Front Mater2022; 9: 851555.
49.
LeeDEChoiYHongG, et al. Parametric image-based concrete defect assessment method. Case Stud Constr Mater2024; 20: e02962.
50.
AkinoshoTDOyedeleLOBilalM, et al. Deep learning in the construction industry: a review of present status and future innovations. J Build Eng2020; 32: 101827.
51.
ZhuD. SIFT algorithm analysis and optimization. In: Proceedings of the 2010 international conference on image analysis and signal processing (IASP), Zhejiang, China, April 2010, 2010, pp. 415–419. Piscataway, NJ: IEEE.
52.
ChengSYangFChenZ, et al. Deep seam prediction for image stitching based on selection consistency loss. arXiv preprint arXiv:2302.05027, 2023.
53.
BradskiG. The OpenCV library. Dr Dobb’s J Softw Tools2000; 25(11): 120, 122–125.
54.
WooSParkJLeeJY, et al. CBAM: convolutional block attention module. In: Proceedings of the European conference on computer vision (ECCV), Munich, Germany, September 8–14, 2018, pp. 3–19. Cham: Springer.
55.
ZhangZSabuncuMR. Generalized cross entropy loss for training deep neural networks with noisy labels. Adv Neural Inf Process Syst2018; 32: 8792–8802.
56.
MilletariFNavabNAhmadiSA. V-Net: fully convolutional neural networks for volumetric medical image segmentation. In: Proceedings of the 2016 4th international conference on 3D vision (3DV), Stanford, CA, USA, October 25–28, 2016, pp. 565–571. Piscataway, NJ: IEEE.
57.
PoudelRPKLiwickiSCipollaR. Fast-SCNN: fast semantic segmentation network. In: Proceedings of the British machine vision conference (BMVC), Cardiff, UK, September 9–12, 2019. London, UK: BMVA Press.
58.
RonnebergerOFischerPBroxT. U-Net: convolutional networks for biomedical image segmentation. In: Medical image computing and computer-assisted intervention – MICCAI 2015, Munich, Germany, October 5–9, 2015, pp. 234–241. Cham: Springer.
59.
ZhaoHShiJQiX, et al. Pyramid scene parsing network. In: Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR), Honolulu, HI, USA, July 21–26, 2017, pp. 6230–6239. Piscataway, NJ: IEEE.
60.
ChenLCZhuYPapandreouG, et al. Encoder–decoder with atrous separable convolution for semantic image segmentation. In: Computer vision – ECCV 2018, Munich, Germany, September 8–14, 2018, pp. 833–851. Cham: Springer.
61.
BadrinarayananVKendallACipollaR. SegNet: a deep convolutional encoder–decoder architecture for image segmentation. IEEE Trans Pattern Anal Mach Intell2017; 39(12): 2481–2495.
62.
XieEWangWYuZ, et al. SegFormer: simple and efficient design for semantic segmentation with transformers. Adv Neural Inf Process Syst2021; 34: 12077–12090.
63.
Jiangsu Provincial Administration for Market Regulation; Department of Housing and Urban-Rural Development of Jiangsu Province. Technical specification for application of fair-faced concrete: DB32/T 4184—2021 (Appendix B: Image acquisition and evaluation method for appearance quality of fair-faced concrete) [Internet]. Nanjing, CN, 2021, https://www.gongbiaoku.com/book/cxy240847h0 (accessed 5 March 2026).
Precast/Prestressed Concrete Institute (PCI). Color and texture selection guide [Internet]. Chicago, IL: PCI, https://www.pci.org/ColorTextureGuide (2003, accessed 5 March 2026).
66.
ParthasarathySSankaranP. An automated multi-scale Retinex with color restoration for image enhancement. In: Proceedings of the 2012 national conference on communications (NCC); Kharagpur, India, February 3–5, 2012, pp. 1–5. Piscataway, NJ: IEEE.
ACI Committee 347. Guide to formed concrete surfaces: ACI PRC-347.3-13(21) [Internet]. Farmington Hills, MI: American Concrete Institute, 2014 (reapproved 2021), https://www.concrete.org/store/productdetail.aspx?ItemID=347313 (2014, reapproved 2021), accessed 5 March 2026).
