深度学习驱动的碳纤维管壁缺陷镜面红外检测无损定量评估方法

杨阳; 陈洪恩; 刘博凡; 仝宗飞; 裴翠祥; 陈振茂

doi:10.7652/xjtuxb202606008

您当前的位置：

首页 >

文章列表页 >

深度学习驱动的碳纤维管壁缺陷镜面红外检测无损定量评估方法

专题　先进无损检测技术 | 更新时间：2026-06-12

- 深度学习驱动的碳纤维管壁缺陷镜面红外检测无损定量评估方法
- Intelligent Quantitative Non-Destructive Evaluation Method for Delamination of Carbon Fiber Composite Tube with Mirror-Reflection Infrared Thermography
- 西安交通大学学报 2026年60卷第6期页码：97-109
- 作者机构：
  
  西安交通大学陕西省无损检测与结构完整性评价工程技术研究中心，710049，西安
  西安交通大学复杂服役环境重大装备结构强度与寿命全国重点实验室，710049，西安
  西安交通大学航天航空学院，710049，西安
- 作者简介：
  
  作者简介：杨阳（1997—），男，博士生；
  仝宗飞（通信作者），男，助理教授。
- 基金信息：
  
  国家自然科学基金资助项目(52477015;12402226)
- DOI：10.7652/xjtuxb202606008
  中图分类号： TP274
- 收稿：2025-09-03，
  
  纸质出版：2026-06-10
- 稿件说明：
移动端阅览
杨阳, 陈洪恩, 刘博凡, 等. 深度学习驱动的碳纤维管壁缺陷镜面红外检测无损定量评估方法[J]. 西安交通大学学报, 2026,60(6):97-109.

YANG Yang, CHEN Hongen, LIU Bofan, et al. Intelligent Quantitative Non-Destructive Evaluation Method for Delamination of Carbon Fiber Composite Tube with Mirror-Reflection Infrared Thermography[J]. Journal of Xi'an Jiaotong University, 2026, 60(6): 97-109.
杨阳, 陈洪恩, 刘博凡, 等. 深度学习驱动的碳纤维管壁缺陷镜面红外检测无损定量评估方法[J]. 西安交通大学学报, 2026,60(6):97-109. DOI： 10.7652/xjtuxb202606008.

YANG Yang, CHEN Hongen, LIU Bofan, et al. Intelligent Quantitative Non-Destructive Evaluation Method for Delamination of Carbon Fiber Composite Tube with Mirror-Reflection Infrared Thermography[J]. Journal of Xi'an Jiaotong University, 2026, 60(6): 97-109. DOI： 10.7652/xjtuxb202606008.

摘要

针对大型可展开天线碳纤维复合材料管壁缺陷传统检测方法存在检测距离近、效率低、智能化不足的问题，提出了一种基于图像处理和深度学习算法的碳纤维管脱粘缺陷镜面激光红外检测智能化定量评估方法。首先，基于Otsu算法进行检测区域自动识别提取，并对提取后的图像序列进行主成分分析，有效抑制噪声干扰并提升缺陷特征的对比度。其次，提出了一种基于深度学习算法的缺陷智能化识别和分割算法，通过决策级融合策略实现缺陷区域的精准快速识别，并基于区域约束分割策略和逻辑或聚合算法实现缺陷形状的准确分割。最后，通过缺陷矫正与三维还原算法，实现了缺陷的三维可视化重构。对含内部脱粘缺陷的碳纤维管试件进行检测，结果表明：采用所提方法得到的缺陷定量评估结果与CT扫描结果高度一致，各个缺陷识别面积的最大误差为7.7%，平均误差为5.1%，均在工程误差允许范围内，验证了所提方法的准确性和有效性。该研究为可展开天线碳纤维管的远距离、高效和智能化无损定量评估提供了一定的指导意义。

Abstract

To address the challenges facing conventional detection methods for delamination of carbon fiber composite tubes of large deployable antennas，namely，short detection range，low efficiency，and insufficient intelligence，an intelligent quantitative evaluation method for delamination of carbon fiber composite tubes with mirror-reflection infrared thermography based on image processing and deep learning algorithms was proposed.First，the detection region was automatically identified and extracted using the Otsu algorithm，and the extracted image sequences were processed by principal component analysis to effectively suppress noise and enhance defect feature contrast.Second，a deep learning-based defect recognition and segmentation algorithm was proposed；defect regions were identified accurately and rapidly through a decision-level fusion strategy，and defect shapes were precisely segmented using a region-constrained segmentation strategy combined with a logical or aggregation algorithm.Finally，defect correction and 3D reconstruction algorithms were applied to achieve 3D visualization of the defects.Carbon fiber composite tube specimens with internal delamination were tested.The results indicated that the quantitative evaluations for defects obtained by the proposed method were in good agreement with computed tomography（CT）results：the maximum error in the identified area of each defect was 7.7%and the mean error was 5.1%，both within allowable engineering tolerances，thereby validating the accuracy and effectiveness of that method.The study is expected to provide guidance for long-range，efficient，andintelligentquantitativenon-destructiveevaluationofdeployableantennacarbonfiber composite tubes.

关键词

Keywords

references

马小飞，李洋，肖勇，等.大型空间可展开天线反射器研究现状与展望[J].空间电子技术，2018，15（2）：16-26.

MA Xiaofei，LI Yang，XIAO Yong，et al.Development and tendency of large space deployable antenna reflector [J].Space Electronic Technology，2018，15（2）：16-26.

BARBERO E J.Introduction to composite materials design[M].2nd ed.Boca Raton，USA：CRC Press，2011.

SMITH R.Composite defects and their detection[J]. Materials Science and Engineering，2009，3（1）：103-143.

TALREJA R，SINGH C V.Damage and failure of composite materials [M].Cambridge，UK：Cambridge University Press，2012.

BOSSI R H，GIURGIUTIU V.Nondestructive testing of damage in aerospace composites [M]//IRVING P E，SOUTIS C.Polymer Composites in the Aerospace Industry. Cambridge，UK：Woodhead Publishing，2015：413-448.

刘青旭，陈海峰，BRYANSKY A，等.航天复合材料结构健康监测技术应用进展[J].复合材料学报，2024，41（9）：4563-4588.

LIU Qingxu，CHEN Haifeng，BRYANSKY A，et al. Progress in application on health monitoring technology for aerospace composite structures[J].Acta Materiae Compositae Sinica，2024，41（9）：4563-4588.

YANG Xiaoyu，VERBOVEN E，JU Bingfeng，et al. Comparative study of ultrasonic techniques for reconstructing the multilayer structure of composites [J]. NDT & E International，2021，121：102460.

魏莹莹，安庆龙，蔡晓江，等.碳纤维复合材料超声扫描分层检测及评价方法[J].航空学报，2016，37（11）：3512-3519.

WEI Yingying，AN Qinglong，CAI Xiaojiang，et al. CFRP ultrasonic scan delamination detection and evalu ation method [J].Acta Aeronautica et Astronautica Sinica，2016，37（11）：3512-3519.

DU Yali，XIE Shejuan，LI Xudong，et al.A fast forward simulation scheme for eddy current testing of crack in a structure of carbon fiber reinforced polymer laminate[J].IEEE Access，2019，7：152278-152288.

GUO Wei，GUO Lihua，XU Hao，et al.Impact damage detection on carbon fiber reinforced polymer tube by a mutual differential Bobbin probe [J].Composites：Part A Applied Science and Manufacturing，2025，193：108806.

傅健，张昌盛，朱国港，等.X射线分层层析成像技术及在航空航天领域的应用[J].航空制造技术，2019，62（14）：49-54.

FU Jian，ZHANG Changsheng，ZHU Guogang，et al. Development and application of X-ray computed laminography for aerospace [J].Aeronautical Manufacturing Technology，2019，62（14）：49-54.

董方旭，王从科，凡丽梅，等.X射线检测技术在复合材料检测中的应用与发展[J].无损检测，2016，38（2）：67-72.

DONG Fangxu，WANG Congke，FAN Limei，et al. The application and development of detection of composite materials by X-ray nondestructive testing techniques [J].Nondestructive Testing，2016，38（2）：67-72.

QIU Jinxing，PEI Cuixiang，LIU Haochen，et al. Quantitative evaluation of surface crack depth with laser spot thermography [J].International Journal of Fatigue，2017，101（Part 1）：80-85.

YANG Ruizhen，HE Yunze.Optically and non-optically excited thermography for composites：a review[J]. Infrared Physics & Technology，2016，75：26-50.

QIU Jinxing，PEI Cuixiang，LIU Haochen，et al.Remote inspection of surface cracks in metallic structures with fiber-guided laser array spots thermography[J]. NDT & E International，2017，92：213-220.

许颖，王青原，罗聪聪，等.基于激光热激励红外热成像纤维增强聚合物复合材料加固混凝土结构界面损伤无损检测[J].复合材料学报，2020，37（2）：472-481.

XU Ying，WANG Qingyuan，LUO Congcong，et al. Nondestructive debonding detection of fiber reinforced plastics strengthened concrete structure based oninfrared thermal imaging with laser thermal excitation[J]. Acta Materiae Compositae Sinica，2020，37（2）：472-481.

KEO S A，DEFER D，BREABAN F，et al.Comparison between microwave infrared thermography and CO 2 laser infrared thermography in defect detection in applications withCFRP [J ] .Materials Sciences and Applications，2013，4（10）：600-605.

KEO S A，BRACHELET F，BREABAN F，et al.Defect detection in CFRP by infrared thermography w ith CO 2 laser excitation compared to conventional lock-in infrared thermography[J ] .Composites：Part B Engineering，2015，69：1-5.

SEMEROK A，JAUBERT F，FOMICHEV S V，et al.Laser lock-in thermography for thermal contact characterisation of surface layer[J].Nuclear Instruments and Methods in Physics Research：Section A Accelerators，Spectrometers，Detectors and Associated Equipment，2012，693：98-103.

LIU Junyan，LIU Yang，WANG Fei，et al.Study on probability of detection（POD）determination using lock-in thermography for nondestructive inspection（NDI）of CFRP composite materials [J].Infrared Physics & Technology，2015，71：448-456.

GONG Jinlong，LIU Junyan，WANG Fei，et al.Inverse heat transfer approach for nondestructive estimation the size and depth of subsurface defects of CFRP composite using lock-in thermography [J].Infrared Physics & Technology，2015，71：439-447.

GAO Bin，BAI Libing，WOO W L，et al.Automatic defect identification of eddy current pulsed thermography using single channel blind source separation[J]. IEEE Transactions on Instrumentation and Measurement，2014，63（4）：913-922.

WANG Zhi，PEI Cuixiang，ZHANG Zhenyu，et al. Quantitative test of delamination defects in CFRP with surface interference by laser thermography[J].Infrared Physics & Technology，2024，136：105046.

SHEPARD S M，LHOTA J R，RUBADEUX B A，et al.Reconstruction and enhancement of active thermographic image sequences [J].Optical Engineering，2003，42（5）：1337-1342.

MALDAGUE X，MARINETTI S.Pulse phase infrared thermography [J].Journal of Applied Physics，1996，79（5）：2694-2698.

RAJIC N.Principal component thermography for flaw contrast enhancement and flaw depth characterization in composite structures [J]. Composite Structures，2002，58（4）：521-528.

LUO Qin，GAO Bin，WOO W L，et al.Temporal and spatial deep learning network for infrared thermal defect detection[J].NDT & EInternational，2019，108：102164.

WU Haiyi，ZHANG Hongwei，HU Guoqing，et al.Deep learning-based reconstruction of the structure of heterogeneous composites from their temperature fields [J]. AIP Advances，2020，10（4）：045037.

WEI Ziang，FERNANDES H，HERRMANN H G，et al.A deep learning method for theimpact damage segmentation of curve-shaped CFRP specimens inspected by infrared thermography[J].Sensors，2021，21（2）：395.

RUAN Lingfeng，GAO Bin，WU Shichun，et al.DeftectNet：joint loss structured deep adversarial network for thermography defect detecting system[J].Neurocomputing，2020，417：441-457.

ALI R，CHA Y J.Attention-based generative adversarial network with internal damage segmentation using thermography[J].Automation in Construction，2022，141：104412.

CHENG Liangliang，TONG Zongfei，XIE Shejuan，et al.IRT-GAN：a generative adversarial network with a multi-headed fusion strategy for automated defect detectionin composites using infrared thermography[J]. Composite Structures，2022，290：115543.

KANG Yukuan，LIU Lei，GAO Bin，et al.Automated thermography cognitive sensing-feedbackinspection for large irregular sample[J].IEEE Transactions on Instrumentation and Measurement，2024，73：1-10.

WU Shunyao，GAO Bin，WOO W L，et al.Defect super-resolution algorithm based on infrared thermal imaging physical kernel [J].NDT & EInternational，2025，154：103368.

WANG Rongbang，PEI Cuixiang，XIA Ruicong，et al.A portable fiber laser thermography system with beam homogenizing for CFRP inspection[J].NDT &E International，2021，124：102550.

TONG Zongfei，CHENG Liangliang，XIE Shejuan，et al.A flexible deep learning framework for thermographic inspection of composites[J].NDT & EInternational，2023，139：102926.

KIRILLOV A，MINTUN E，RAVI N，et al.Segment anything[C]//2023 IEEE/CVF International Conference on Computer Vision（ICCV）.Piscataway，NJ，USA：IEEE，2023：3992-4003.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

碳纤维管壁缺陷镜面红外检测的智能无损定量评估方法研究

面向模型预测控制的风电机组深度Koopman全局线性建模方法

主蒸汽管道破裂事故下蒸汽发生器的动态液位预测

融合U-net网络的纯卷积视频预测模型

一种双阻尼小波赋能的可解释卷积神经网络在轴承故障诊断中的应用