基于深度学习的外墙饰面砖系统空鼓定量识别与评估
PDF下载 (89)潘金晶,童璐璐,娄添添,章子华,叶春阳.基于深度学习的外墙饰面砖系统空鼓定量识别与评估[J].宁波大学学报(理工版),2026,39(1):26-35.DOI:10.20098/j.cnki.1001-5132.2025.0130
PAN Jinjing,TONG Lulu,LOU Tiantian,ZHANG Zihua,YE Chunyang.Quantitative identification and evaluation of
hollowing in exterior wall tiling system based on deep learning
[J].Journal of Ningbo University(Natural Science & Engineering Edition),2026,39(1):26-35.DOI:10.20098/j.cnki.1001-5132.2025.0130
| Title: | Quantitative identification and evaluation of
hollowing in exterior wall tiling system based on deep learning |
| 作者: | 潘金晶, 童璐璐, 娄添添, 章子华, 叶春阳 |
| Author(s): | PAN Jinjing, TONG Lulu, LOU Tiantian, ZHANG Zihua, YE Chunyang |
| 关键词: | 饰面砖系统; 空鼓; 红外热像; 深度学习; 实例分割; 定量评估 |
| Keywords: | tiling systems; tile hollowing; infrared thermography; deep learning; instance segmentation; quantitative evaluation |
| 分类号: | TU767 |
| DOI: | 10.20098/j.cnki.1001-5132.2025.0130 |
| 文献标识码: | A |
| 摘要: | 为实现外墙饰面砖系统空鼓的自动识别与定量评估，采用红外相机拍摄并建立包含1686个实例的图像数据库，然后根据真实几何关系对饰面砖和缺陷进行标定，用以评估深度学习实例分割方法中Mask R-CNN与YOLACT模型在空鼓及饰面砖识别上的性能与效率，并基于Python自主开发饰面砖空鼓率计算与可视化程序。结果表明，Mask R-CNN与YOLACT模型在数据集上均表现良好，平均分割精度分别达到0.855和0.771，其中YOLACT模型在计算效率与模型复杂度方面表现更优。自主开发程序能直观、量化反映空鼓区域及所占比例，可有效辅助准确判断缺陷程度。 |
| Abstract: | To achieve automatic identification and quantitative evaluation of hollowing in exterior wall tiling system, an infrared camera was used to capture hollowing images, with which an image dataset containing 1686 instances was established and labeled based on real geometric relationships. The performance and efficiency of the Mask R-CNN and YOLACT models in deep learning-based instance segmentation were evaluated for hollowing and tiles detection, and a Python-based program was developed to calculate and visualize the hollowing rate of tiles. The results show that Mask R-CNN and YOLACT models perform well on the dataset, with mAP being 0.855 and 0.771, respectively. The YOLACT model demonstrates superior computational efficiency and model complexity. The developed program can effectively and intuitively visualize the hollowing areas and proportions, providing valuable assistance in accurately evaluating the defect severity in exterior wall tiling system. |
| 参考文献 /References: | [1].冯力强, 王欢祥, 晏大玮, 等. 红外热像法检测建筑外墙饰面层内部缺陷试验研究[J]. 土木工程学报, 2014, 47(6):51-56. [2].PEREIRA C, DE BRITO J, SILVESTRE J D. Contribution of humidity to the degradation of facade claddings in current buildings[J]. Engineering failure analysis, 2018, 90:103-115. [3].刘强, 胡玉琨, 朱红光, 等. 基于红外检测的外墙饰面砖粘结[J]. 建筑技术, 2015, 46(10):924-926. [4].庞森, 王长军, 王健, 等. 工程渗漏检测技术研究进展及展望[J]. 工业建筑, 2023, 53(增刊2):883-888. [5].BARBOSA M T G, ROSSE V J, LAURINDO N G. Thermography evaluation strategy proposal due moisture damage on building facades[J]. Journal of building engineering, 2021, 43:102555. [6].张俊傥, 史夏杰, 李光明, 等. 红外热成像技术在古建安全评价中的应用[J]. 工业建筑, 2014(增刊1):219- 222, 192. [7].HATIR M E, INCE I, BOZKURT F. Investigation of the effect of microclimatic environment in historical buildings via infrared thermography[J]. Journal of building engineering, 2022, 57:104916. [8].GARRIDO I, BARREIRA E, ALMEIDA R M S F, et al. Introduction of active thermography and automatic defect segmentation in the thermographic inspection of specimens of ceramic tiling for building facades[J]. Infrared physics & technology, 2022, 121:104012. [9].LAI W W L, LEE K K, POON C S. Validation of size estimation of debonds in external wall’s composite finishes via passive infrared thermography and a gradient algorithm[J]. Construction and building materials, 2015, 87(15):113-124. [10].LI Q X, PENG X, ZHONG X G, et al. Quantitative identification of debonding defects in building facades based on UAV-thermography using a two-stage network integrating dual attention mechanism[J]. Infrared physics & technology, 2024, 138:105241. [11].CAO M T. Drone-assisted segmentation of tile peeling on building facades using a deep learning model[J]. Journal of building engineering, 2023, 80:108063. [12].米庆仁. 基于深度学习多模态RGB-T图像融合的建筑外墙缺陷检测研究[D]. 哈尔滨: 哈尔滨工业大学, 2024. [13].刘颖. 基于深度学习的红外墙体裂缝高精度检测技术研究[D]. 北京: 北京邮电大学, 2021. [14].蔡茹莹. 基于计算机视觉的建筑物外墙裂缝智能检测及其几何量化评级研究[D]. 深圳: 深圳大学, 2022. [15].MELO A C, COSTA E SILVA A J, TORRES S M, et al. Influence of the contact area in the adherence of mortar: ceramic tiles interface[J]. Construction and building materials, 2020, 243:118274. [16].GARRIDO I, LAGÜELA S, OTERO R, et al. Thermographic methodologies used in infrastructure inspection: a review-post-processing procedures[J]. Applied energy, 2020, 266:114857. [17].CURCI H C F, DE ANDRADE R P, MARANHÃO F L, et al. Analysis of adhered tiling systems based on experimental evaluation and numerical modeling[J]. Construction and building materials, 2022, 325:126746. [18].WINNEFELD F, KAUFMANN J, HACK E, et al. Moisture induced length changes of tile adhesive mortars and their impact on adhesion strength[J]. Construction and building materials, 2012, 30:426-438. [19].中华人民共和国住房和城乡建设部. 红外热像法检测建筑外墙饰面粘结质量技术规程: JGJ/T 277-2012[S]. 北京: 中国建筑工业出版社, 2012. [20].RUSSELL B C, TORRALBA A, MURPHY K P, et al. LabelMe: a database and web-based tool for image annotation[J]. International journal of computer vision, 2008, 77(1):157-173. [21].BOLYA D, ZHOU C, XIAO F Y, et al. Yolact: real-time instance segmentation[C]//IEEE/CVF International Conference on Computer Vision (ICCV). Seoul, South Korea. 2019:9156-9165. [22].HE K M, GKIOXARI G, DOLLÁR P, et al. Mask r-CNN[C]//IEEE International Conference on Computer Vision and Pattern Recongnition. Hawaii, USA. 2017: 2961-2969. [23].REN S, HE K, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE transactions on pattern analysis and machine intelligence, 2016, 39(6):1137-1149. [24].HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]//IEEE International Conference on Computer Vision and Pattern Recognition. Las Vegas, USA. 2016:770-778. [25].LIN T Y, DOLLÁR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]//IEEE International Conference on Computer Vision and Pattern Recognition. Hawaii, USA. 2017:2117-2125. [26].DENG J, DONG W, SOCHER R, et al. Imagenet: a large-scale hierarchical image database[C]//IEEE Inter- national Conference on Computer Vision and Pattern Recognition. Miami, USA. 2009:248-255. [27].CHEN K, WANG J Q, PANG J M, et al. MMDetection: open MMlab detection toolbox and benchmark[EB/OL]. [2024-03-21]. https://arxiv.org/abs/1906.07155v1. [28].LIN T Y, MAIRE M, BELONGIE S, et al. Microsoft COCO: common objects in context[C]//13th European Conference on Computer Vision (ECCV). Zurich, Switzerland. 2014:740-755. |
| 备注/Memo: | 收稿日期：2025-01-24 宁波大学学报（理工版）网址：http://journallg.nbu.edu.cn/ 基金项目：国家自然科学基金（52378162）；宁波市“科创甬江2035”重大应用示范项目（2024Z014） 第一作者：潘金晶，硕士研究生，主要研究方向为外墙饰面层病害检测与加固。E-mail: panjinjing1208@163.com *通信作者：章子华，博导/教授，主要研究方向为基础设施病害诊治理论与技术。E-mail: zhangzihua@nbu.edu.cn 宁波大学学报（理工版）网址：http://journallg.nbu.edu.cn/ |
