3D-printed concrete apparent defect detection method based on improved YOLOv8

doi:10.3785/j.issn.1008-973X.2026.04.015

Journal of ZheJiang University (Engineering Science)

2026, Vol. 60

Issue (4): 833-843 DOI: 10.3785/j.issn.1008-973X.2026.04.015

3D-printed concrete apparent defect detection method based on improved YOLOv8

Wei TIAN(

),Linhong ZHOU,Xinyang LI,Jianming WANG,Yukang HUANG

School of Civil Engineering, Xi’an University of Architecture and Technology, Xi’an 710055, China

Download:

HTML

PDF(2032KB) HTML
Export: BibTeX | EndNote (RIS)

Abstract

An improved YOLOv8-SRA-based method was developed for apparent defect detection in 3D-printed concrete to improve the quality control level of 3D-printed concrete. Real-time detection challenges and low detection accuracy were addressed for defects including holes, pocked surfaces, fractures, and collapses. Initially, the Small head module was inserted into the Head layer with a view to enhancing the detection of minor defects. Concurrently, the Restormer attention mechanism module was incorporated into the Backbone layer, with the objective of enhancing the extraction of multi-scale defect features in complex backgrounds. Furthermore, the AKConv module was incorporated into the Neck and Backbone layers with a view to enhancing the extraction and detection of irregular defects. Ultimately, the synergistic effectiveness of the modules and the model’s superiority were confirmed by means of ablation experiments and comparative tests. The modified model achieved a 94.7% mAP@0.5, a 92.6% recall rate, and a 91.3% precision rate, which represented enhancements of 6.2, 9.7, and 7.4 percentage points, respectively, when compared to the original model. These enhancements addressed the limitations of the original model, such as its tendency to overlook minor defects and its proclivity to erroneously detect irregular defects. This study offers a novel approach to the quality control of 3D-printed concrete buildings and structures in practical engineering applications.

Key words： concrete engineering 3D-printed concrete apparent defect detection quality control YOLOv8 model

Received: 04 July 2025 Published: 19 March 2026

CLC:	TU 528
	TP 399

	Service
	E-mail this article
	Add to my bookshelf
	Add to citation manager
	E-mail Alert
	RSS
	Articles by authors
	Wei TIAN
	Linhong ZHOU
	Xinyang LI
	Jianming WANG
	Yukang HUANG

Cite this article:

Wei TIAN,Linhong ZHOU,Xinyang LI,Jianming WANG,Yukang HUANG. 3D-printed concrete apparent defect detection method based on improved YOLOv8. Journal of ZheJiang University (Engineering Science), 2026, 60(4): 833-843.

URL:

https://www.zjujournals.com/eng/10.3785/j.issn.1008-973X.2026.04.015 OR https://www.zjujournals.com/eng/Y2026/V60/I4/833

基于改进YOLOv8的3D打印混凝土表观缺陷检测方法

为了提升3D打印混凝土质量控制水平，针对孔洞、麻面、断裂、塌陷等表观缺陷实时检测难、检测精度低的问题，提出基于改进YOLOv8-SRA的3D打印混凝土表观缺陷检测方法. 在Head层中插入Small head模块，提高对微小缺陷的检测能力；在Backbone层添加Restormer注意力机制模块，增强在复杂背景下对多尺度缺陷特征的提取能力；在Neck层与Backbone层上添加AKConv模块，提高对于不规则缺陷的特征提取与检测能力. 通过消融实验与对比试验验证各模块的协同有效性与模型的优越性. 研究表明：改进后模型mAP@0.5为94.7%、召回率为92.6%、精确率为91.3%，较原模型分别提升了6.2、9.7、7.4个百分点，解决了原模型对微小缺陷易漏检、对不规则缺陷易误检的问题，为实际工程中3D打印混凝土建筑及结构质量控制提供了新思路.

关键词： 混凝土工程, 3D打印混凝土, 表观缺陷检测, 质量控制, YOLOv8模型

Tab.1 3D-printed concrete simulation test equipment

Fig.1 Diagram of shooting locations and viewpoints

Fig.2 YOLOv8 with small target detection head added

Fig.3 Structure of Restormer’s attention mechanism

Fig.4 Structure of MDTA module

Fig.5 Structure of GDFN module

Fig.6 Structure of AKConv module

Tab.2 Computer network environment

Fig.7 Representation of apparent defects in 3D printed concrete

Tab.3 Analysis of YOLOv8 model detection results before and after improvement

Fig.8 Comparison of surface defect detection effects for 3D-printed concrete before and after improvement

Fig.9 Precision versus recall curve comparison before and after improvement

Tab.4 Results of ablation experiments on each module of YOLOv8-SRA model

Fig.10 Comparison of thermal values among different modules

Tab.5 Comparative experimental results for each model

Fig.11 Comparison of detection results of different models


[1]	刘化威, 刘超, 白国良, 等基于孔结构缺陷的3D打印粗骨料混凝土力学性能试验研究[J]. 土木工程学报, 2022, 55 (12): 54- 64 LIU Huawei, LIU Chao, BAI Guoliang, et al Experimental study on mechanical properties of 3D printed coarse aggregate concrete based on the pore structure defects[J]. China Civil Engineering Journal, 2022, 55 (12): 54- 64

[2]	SHAKOR P, NEJADI S, PAUL G, et al Review of emerging additive manufacturing technologies in 3D printing of cementitious materials in the construction industry[J]. Frontiers in Built Environment, 2019, 4: 85 doi: 10.3389/fbuil.2018.00085

[3]	史庆轩, 万胜木 3D打印混凝土工作及力学性能研究进展[J]. 工业建筑, 2022, 52 (5): 208- 218 SHI Qingxuan, WAN Shengmu Research progress on working and mechanical properties of 3D printed concrete[J]. Industrial Construction, 2022, 52 (5): 208- 218

[4]	PANDA B, PAUL S C, MOHAMED N A N, et al Measurement of tensile bond strength of 3D printed geopolymer mortar[J]. Measurement, 2018, 113: 108- 116 doi: 10.1016/j.measurement.2017.08.051

[5]	BUSWELL R A, LEAL DE SILVA W R, JONES S Z, et al 3D printing using concrete extrusion: a roadmap for research[J]. Cement and Concrete Research, 2018, 112: 37- 49 doi: 10.1016/j.cemconres.2018.05.006

[6]	JI G, DING T, XIAO J, et al A 3D printed ready-mixed concrete power distribution substation: materials and construction technology[J]. Materials, 2019, 12 (9): 1540 doi: 10.3390/ma12091540

[7]	CUI H, YU S, CAO X, et al Evaluation of printability and thermal properties of 3D printed concrete mixed with phase change materials[J]. Energies, 2022, 15 (6): 1978 doi: 10.3390/en15061978

[8]	吴子燕, 贾大卫, 王其昂基于卷积神经网络与区域生长法的建筑裂缝识别[J]. 应用基础与工程科学学报, 2022, 30 (2): 317- 327 WU Ziyan, JIA Dawei, WANG Qi’ang Building crack identification based on convolutional neural network and regional growth method[J]. Journal of Basic Science and Engineering, 2022, 30 (2): 317- 327

[9]	XU G, HAN X, ZHANG Y, et al Dam crack image detection model on feature enhancement and attention mechanism[J]. Water, 2023, 15 (1): 64 doi: 10.3390/w15010064

[10]	余加勇, 李锋, 薛现凯, 等基于无人机及Mask R-CNN的桥梁结构裂缝智能识别[J]. 中国公路学报, 2021, 34 (12): 80- 90 YU Jiayong, LI Feng, XUE Xiankai, et al Intelligent identification of bridge structural cracks based on unmanned aerial vehicle and mask R-CNN[J]. China Journal of Highway and Transport, 2021, 34 (12): 80- 90

[11]	罗大明, 谢俊科, 李凡, 等. 基于改进YOLOv5模型的混凝土结构表观损伤检测方法[EB/OL]. [2025−04−11]. https://link.cnki.net/doi/10.15986/j.1006-7930.2025.02.001.

[12]	曾妮, 马宗方, 宋琳, 等 3D打印混凝土界面孔隙智能检测方法研究[J]. 工程设计学报, 2025, 32 (1): 11- 22 ZENG Ni, MA Zongfang, SONG Lin, et al Research on intelligent detection method of 3D printed concrete interface pore[J]. Chinese Journal of Engineering Design, 2025, 32 (1): 11- 22

[13]	李金沛, 孟晓林, 胡亮亮, 等基于改进YOLOv8的桥梁小目标裂缝检测[J]. 清华大学学报: 自然科学版, 2025, 65 (7): 1260- 1271 LI Jinpei, MENG Xiaolin, HU Liangliang, et al Bridge small target crack detection based on improved YOLOv8[J]. Journal of Tsinghua University: Science and Technology, 2025, 65 (7): 1260- 1271

[14]	董红召, 林少轩, 佘翊妮交通目标YOLO检测技术的研究进展[J]. 浙江大学学报:工学版, 2025, 59 (2): 249- 260 DONG Hongzhao, LIN Shaoxuan, SHE Yini Research progress of YOLO detection technology for traffic object[J]. Journal of Zhejiang University: Engineering Science, 2025, 59 (2): 249- 260

[15]	武晓春, 张恒骏, 谭磊基于YOLOv8-HSV的隧道螺栓锈蚀检测及等级判定[J]. 浙江大学学报:工学版, 2025, 59 (10): 2144- 2153+2220 WU Xiaochun, ZHANG Hengjun, TAN Lei Corrosion detection and grade determination of tunnel bolts based on YOLOv8-HSV[J]. Journal of Zhejiang University: Engineering Science, 2025, 59 (10): 2144- 2153+2220

[16]	刘鑫蕊, 朱树先. 基于YOLOv8改进的复杂场景目标检测 [EB/OL]. (2025−04−08). https://kns.cnki.net/kcms/detail/61.1123.tn.20250408.1027.014.html.

[17]	付浩辰, 白堂博, 许贵阳, 等基于改进YOLOv8的轨道板裂缝检测算法[J]. 北京交通大学学报, 2024, 48 (6): 133- 143 FU Haochen, BAI Tangbo, XU Guiyang, et al Crack detection in track slabs based on an improved YOLOv8 algorithm[J]. Journal of Beijing Jiaotong University, 2024, 48 (6): 133- 143

[18]	ELSHARKAWY Z F, KASBAN H, ABBASS M Y Efficient surface crack segmentation for industrial and civil applications based on an enhanced YOLOv8 model[J]. Journal of Big Data, 2025, 12 (1): 16 doi: 10.1186/s40537-025-01065-1

[19]	AL-TAMIMI A K, ALQAMISH H H, KHALDOUNE A, et al Framework of 3D concrete printing potential and challenges[J]. Buildings, 2023, 13 (3): 827 doi: 10.3390/buildings13030827

[20]	ZAMIR S W, ARORA A, KHAN S, et al. Restormer: efficient transformer for high-resolution image restoration [C]// IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans: IEEE, 2022: 5718–5729.

[21]	王皓晨, 刘灵婧, 凌怡, 等 AKConv模块驱动的YOLOv8轻量化目标检测模型[J]. 计算机时代, 2025, (3): 50- 54 WANG Haochen, LIU Lingjing, LING Yi, et al AKConv-driven lightweight YOLOv8 object detection model[J]. Computer Era, 2025, (3): 50- 54

[22]	ZHANG X, SONG Y, SONG T, et al. AKConv: convolutional kernel with arbitrary sampled shapes and arbitrary number of parameters [EB/OL] (2023–11–20). https://arxiv.org/abs/2311.11587v1.

[23]	HE T, ZHANG Z, ZHANG H, et al. Bag of tricks for image classification with convolutional neural networks [C]// IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach: IEEE, 2019: 558–567.

[24]	GIRSHICK R. Fast R-CNN [C]// IEEE International Conference on Computer Vision. Santiago: IEEE, 2015: 1440–1448.

[1]	HU Jing, QI Guo-Ning, YANG Jing-Hai, et al. Research and implementation of quality control oriented to mass customization[J]. Journal of ZheJiang University (Engineering Science), 2009, 43(12): 2293-2299.

Viewed

Full text

Abstract

Cited

Shared

Discussed