[1]杨 晨,何海风,陈 奎,等.基于YOLOv11n改进的SLM 金属内部缺陷检测方法研究[J].机械与电子,2026,44(06):96-104.
　YANG Chen,HE Haifeng,CHEN Kui,et al.An Improved YOLOv11n based Method for Internal Defect Detection in SLM Metal Additive Manufactured Metals[J].Machinery & Electronics,2026,44(06):96-104.
点击复制

基于YOLOv11n改进的SLM 金属内部缺陷检测方法研究()
分享到：

《机械与电子》[ISSN:1001-2257/CN:52-1052/TH]

卷:
44
期数:
2026年06期
页码:
96-104
栏目:
出版日期:
2026-06-27

文章信息/Info

Title:
An Improved YOLOv11n based Method for Internal Defect Detection in SLM Metal Additive Manufactured Metals
文章编号:
1001-2257(2026)06-0096-09
作者:
杨 晨1; 何海风1; 陈 奎2; 陈华伟3; 郭永力1; 昌远康1
1.贵州师范大学机械与电气工程学院,贵州 贵阳 550025;
2.重庆宇红轨道车辆配件有限公司,重庆 401420;
3.广安理工学院机械与电气工程学院,四川 广安 638000
Author(s):
YANG Chen1; HE Haifeng1; CHEN Kui2; CHEN Huawei3; GUO Yongli1; CHANG Yuankang1
(1.School of Mechanical and Electrical Engineering,Guizhou Normal University,Guiyang 550025,China;
2.Chongqing Yuhong Rail Vehicle Parts Co.,Ltd.,Chongqing 401420,China;
3.School of Mechanical and Electrical Engineering,Guang’an Institute of Technology,Guang’an 638000,China)
关键词:
选择性激光熔融; 缺陷检测; SobelEdgeFusion边缘检测; 轻量化检测头
Keywords:
selective laser melting; defect detection; SobelEdgeFusionedge detection; lightweight detectionhead
分类号:
TP391.4
文献标志码:
A
摘要:
针对SLM 增材制造金属内部缺陷检测问题,提出一种基于YOLOv11n的改进算法SELD-YOLOv11n。首先,设计边缘检测模块SobelEdgeFusion,对缺陷区域进行边缘增强并与主干网络融合,提升小目标缺陷的识别能力;其次,引入轻量化的C2f Faster模块,在保证特征表达能力的同时降低参数量;最后,提出轻量化检测头LD Head及其配套损失函数LD loss,进一步减少计算量并维持模型的稳定性与判别能力。实验结果表明,与基准模型YOLOv11n相比,SELD YOLOv11n的平均精度(mAP)提高3.8百分点,精确率(P)和召回率(R)分别提高6.1百分点和9.7百分点,参数量(Params)减少11.5%,计算复杂度降低10.6%。该方法有效提升了基于深度学习的SLM 增材制造金属内部缺陷检测的准确率。
Abstract:
To address the challenge of detecting internal defects in metals produced by selective laser melting (SLM) additive manufacturing,an improved algorithm based on YOLOv11n,named SELD-YOLOv11n,is proposed.First,an edge detection module (SobelEdgeFusion) is designed to enhance the edges of defect regionsand fuse them with the backbone network,thereby improving the recognition capability for small defects.Second,a lightweight C2f-Faster module is introduced to reduce the number of parameters while maintaining strong feature representation capability.Finally,a lightweighting detection head (LD-Head),along with its corresponding loss function (LD-Loss),is developed to further reduce computational cost while ensuring model stability and discrimination power.Experimental results show that,compared with the baseline YOLOv11n model,SELD-YOLOv11n improves the mean Average Precision (mAP) by 3.8 percentage points,and increases the precision (P) and recall rate (R) by 6.1 and 9.7 percentage points,respectively.Meanwhile,the number of parameters is reduced by 11.5%,and the computational complexity drops by 10.6%.The proposed method effectively enhances the accuracy of deep learning based internal defect detection in SLM additive manufactured metals.

参考文献/References:

[1] Lee K H,Yun G J.A novel heat source model for analysisof melt Pool evolution in selective laser melting process[J].Additive Manufacturing,2020,36:101497.

[2] Wang Xiebin,Yu Jingya,Liu Jiangwei,et al.Effect of process parameters on the phase transformation behavior and tensile properties of NiTi shape memory alloys fabricated by selective laser melting[J].Additive Manufacturing,2020,36:101545.[3] Kruth J P,Levy G,Klocke F,et al.Consolidation phenomena in laser and powder-bed based layered manufacturing[ J].CIRP Annals Manufacturing Technology,2007,56(2):730-759.
[4] Yu Banglong,Wang Ping,Zhao Peng,et al.Impact of defects on high cycle fatigue life in wire-arc additive manufactured TC17 alloy[J].Engineering Failure Analysis,2025,174:109480.
[5] W?glewski W,Bochenek K,Basista M,et al.Comparative assessment of Young’s modulus measurements of metal– ceramic composites using mechanical and non-destructive tests and micro-CT based computational modeling[J].Computational Materials Science,2013,77:19-30.
[6] Thompson A,Maskery I,Leach K R.X-ray computed tomography for additive manufacturing:a review[J].Measurement Science and Technology,2016,27(7):072001.
[7] 王伟,孙轩,张丽娟,等.基于深度学习人工特征混合模型的电弧熔丝增材制造缺陷在线识别[J].现代制造工程,2025(11):116-123.
[8] 王彪,肖华强.激光熔化沉积FeCoNiCrMn高熵合金微观组织及其性能[J/OL].激光技术,2025-11-28.https:∥ link.cnki.net/urlid/51.1125.TN.20251128.1416.005.
[9] Zhou Xiangman,Zheng Shicheng,Li Runsheng,et al. Improved YOLOv5-based pore defect detection algorithm for wire arc additive manufacturing[J].Materials Today Communications,2024,39:108710.
[10] Choi B,Choi Y,Lee M G,et al.Defect detection using deep learning-based YOLOv3 in cross-sectional image of additive manufacturing[J].Archives of Metallurgy and Materials,2021:1037-1041.
[11] Wang Wei,Wang Peiren,Zhang Hanzhong,et al.A real-time defect detection strategy for additive manufacturing processes based on deep learning and machine vision technologies [J].Micromachines,2023,15(1):28.
[12] Huang Yunli,Zhou Xiangman,Wang Guilan,et al.Lightweight defect detection algorithm for wire and arc additive manufacturing based on modified YOLOv8 model[J].Journal of Real-Time Image Processing,2025,22(3):129.
[13] Zubayer H M,Zhang Chaoqun,Wang Yafei.Deep learning-based automatic defect detection of additive manufactured stainless steel [J].Metals,2023,13(12):1987.
[14] 胡彬,朱文彬,王鸣昕,等.基于YOLOv10n的BGA锡球缺陷检测算法[J].半导体技术,2025,50(10):1067-1077.
[15] Li A,Hamzah R,Rahim S K N A,et al.YOLO algorithm with hybrid attention feature pyramid network for solder joint defect detection[J].IEEE Transactions on Components,Packaging and Manufacturing Technology,2024,14(8):1493-1500.
[16] Khanam R,Hussain M.YOLOv11:an overview of the key architectural enhancements [PP/OL].V1.arXiv (2024-10-23) [2026-03-08].https:∥arxiv.org/abs/2410.17725.
[17] Chandra N,Vaidya H,Sawant S,et al.Anovel attention-based generalized efficient layer aggregation network for landslide detection from satellite data inthe higher himalayas,nepal[J].Remote Sensing,2024,16(14):2598.
[18] Sheng Tao,Feng Chen,Zhuo Shaojie,et al.A quantization-friendly separable convolution for mobilenets [C]∥ 2018 1st Workshop on Energy Efficient MachineLearning and Cognitive Computing for Embedded Applications (EMC2).New York:IEEE,2018:14-18.
[19] 曹鸿博,安维胜,梁海鹏等.基于边缘特征增强的 YOLOv8n道路缺陷检测算法[J].电子测量技术,2025,48(19):183-192.
[20] Zhu Jin,Hu Tao,Zheng Linhan,et al.YOLOv8-C2f Faster-EMA:an improved underwater trash detection model based on YOLOv8[J].Sensors,2024,24(8):2483.
[21] Ren Xutong,Xie Lingxi,Wei Chen,et al.Generalized coarse-to-fine visual recognition with progressive training [PP/OL].V2.arXiv (2019-04-16) [2026-03-08].https:∥arxiv.org/abs/1811.12047v2.

[22] Chollet F.Xception:deep learning with depthwise separable convolutions[C]∥ 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). New York:IEEE,2017:1251-1258.
[23] Rumelhart D E,Hinton G E,Williams R J.Learning representations by back-propagating errors[J].Nature, 1986,323:533-536.
[24] Lang Yao,Zheng Dan.An improved sobel edge detection operator[C]∥2016 6th International Conference on Mechatronics,Computer and Education Informationization (MCEI 2016).Paris:Atlantis Press,2016:590-593.
[25] Yang Yang,Wei Lianxin.Grey relevancy degree and improved eight-direction sobel operator edge detection[ J].Journal of Signal and Information Processing,2021,12:43-55.
[26] Ioffe S,Szegedy C.Batch normalization:accelerating deep network training by reducing internal covariate shift[C]∥ ICML’15:Proceedings of the 32nd International Conference on Machine Learning,2015,37:448-456.
[27] Nair V,Hinton G E.Rectified linear units improve restricted boltzmann machines[C]∥ ICML’10:Proceedings of the 27th International Conference on International Conference on Machine Learning,2010:807-814.
[28] Hu Jie,Shen Li,Sun Guang.Squeeze-and-excitation networks[C]∥ 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.New York: IEEE,2018:7132-7141.
[29] Lin Min,Chen Qiang,Yan Shuicheng.Network in network[PP/OL].V2.arXiv(2013-09-18) [2026-03-08].https:∥arxiv.org/abs/1312.4400.
[30] Shannon C E.A mathematical theory of communication[ J].The BellSystem Technical Journal,1948,27(3):379-423.
[31] Luengo J,Moreno R,Sevillano I,et al.A tutorial on the segmentation of metallographic images:taxonomy, new MetalDAM dataset,deep learning-based ensemble model,experimental analysis and challenges [J].Information Fusion,2022,78:232-253.
[32] 井方科,任红格,李松.基于多尺度特征融合的小目标交通标志检测[J].激光与光电子学进展,2024,61(12):372-380.
[33] Zhou Kaiyang,Lei Dong,Chun Pangjo,et al.Evaluation of BFRP strengthening and repairing effects on concrete beams using DIC and YOLO-v5 object detection algorithm[J].Construction and Building Materials,2024,411:134594.
[34] Gupta C,Gill N S,Gulia P,et al.A novel finetuned YOLOv6 transfer learning model for real-time object detection[J].Journal of Real-Time Image Processing,2023,20(3):42.
[35] Varghese R,Sambath M.YOLOv8:A novel object detection algorithm with enhanced performance and robustness[C]∥2024 International Conference on Advances in Data Engineering and Intelligent Computing Systems (ADICS).New York:IEEE,2024:1-6.
[36] Wang C Y,Yeh I H,Liao H Y M.YOLOv9:Learning what you want to learn using programmable gradient information[C]∥ Computer Vision – ECCV 2024.Cham:Springer Nature Switzerland,2024:1 21.
[37] Wang Ao,Chen Hui,Liu Lihao,et al.YOLOv10:Real-time end-to-end object detection[C]∥Advances in Neural Information Processing Systems 37 (NeurIPS2024),2024,37:107984-108011.
[38] Sun Xudong,Wu Pengcheng,Hoi S C H.Face detection using deep learning:An improved faster RCNN approach[J].Neurocomputing,2018,299:42-50.
[39] Zhang Yin,Ye Mu,Zhu Guiyi,et al.FFCA-YOLO for small object detection in remote sensing images[J].IEEE Transactions on Geoscience and Remote Sensing,2024,62:3363057.
[40] Xu Canyang,Liao Yingying,Liu Yongqiang,et al. Lightweight rail surface defect detection algorithm based on an improved YOLOv8[J].Measurement, 2025,242:115922.
[41] Liu Yuan,Liu Yilong,Guo Xiaoyan,et al.Metal surface defect detection using SLF-YOLO enhanced YOLOv8 model[J] .Scientific Reports,2025,15:11105.

相似文献/References:

[1]陆 晓,高 超,蒋承伶,等.基于双注意力机制和级联检测框架的金具螺栓缺陷检测研究[J].机械与电子,2024,42(01):63.
　LU Xiao,GAO Chao,JIANG Chengling,et al.Research on Bolt Defect Detection Based on Dual-attention Mechanism and Cascade Detection Framework[J].Machinery & Electronics,2024,42(06):63.
[2]赖肖强,胡 泓.基于机器视觉算法和改进 MobileNetV2 的晶圆缺陷检测研究[J].机械与电子,2025,(01):34.
　LAI Xiaoqiang,HU Hong.Research on Wafer Defect Detection Based on Machine Vision Algorithms and Improved MobileNetV2[J].Machinery & Electronics,2025,(06):34.
[3]黄心玥,王明泉,耿宇杰,等.基于 YOLOv8 的铝合金铸造轮毂缺陷检测算法研究与分析[J].机械与电子,2025,(02):26.
　HUANG Xinyue,WANG Mingquan,GENG Yujie,et al.Research and Analysis of Defect Detection Algorithm for Aluminum Alloy Casting Wheels Based on YOLOv8[J].Machinery & Electronics,2025,(06):26.
[4]曹振锋,王明泉,路宇鹏,等.基于 YOLOv8 的轻量化轮毂缺陷检测算法研究[J].机械与电子,2025,(06):31.
　CAO Zhenfeng,WANG Mingquan,LU Yupeng,et al.Research on the Lightweight Wheel Hub Defect Detection Algorithm Based on YOLOv8[J].Machinery & Electronics,2025,(06):31.
[5]刘 骞,陈茂林.基于深度学习的少样本光伏边框划痕检测[J].机械与电子,2025,(08):47.
　LIU Qian,CHEN Maolin.Few-shot Scratch Detection on Photovoltaic Frame Based on Deep Learning[J].Machinery & Electronics,2025,(06):47.
[6]王立杰,高嘉伟,徐相龙,等. 基于SDA-YOLOv8的PCBA 缺陷检测算法研究[J].机械与电子,2026,44(01):45.
　WANG Lijie,GAO Jiawei,XU Xianglong,et al. Research on PCBA Defect Detection Algorithm Based on SDA -YOLOv8[J].Machinery & Electronics,2026,44(06):45.
[7]张晓君,陈 平,刘 宾.基于 YOLOv12 的轻量化带钢板表面缺陷检测算法研究[J].机械与电子,2026,44(05):49.
　ZHANG Xiaojun,CHEN Ping,LIU Bin.Research on Lightweight Surface Defect Detection Algorithm Based on YOLOv12 for Strip Steel[J].Machinery & Electronics,2026,44(06):49.

备注/Memo

备注/Memo:
收稿日期:2026-03-26
基金项目:贵州省科技厅项目(黔科合基础MS[2025]248);重庆市技术创新与应用发展专项(CSTB2025TIAD-qykjggX0454)
作者简介:杨 晨 (1999-),男,江西上饶人,硕士研究生,研究方向为深度学习及图像识别;何海风 (1994-),男,贵州遵义人,博士,副教授,研究方向为智能制造,通信作者,E-mail:haifenghe@gznu.edu.cn。
更新日期/Last Update: 2026-08-27