The quality of concrete directly impacts the safety and durability of modern infrastructure. Traditional manual assessment relies on empirical judgment, which is highly subjective and lacks consistency. A concrete defect detection method is proposed based on gradient boosting decision tree (GBDT) and laser-induced acoustic techniques. The wavelet packet decomposition (WPD) method is employed to decompose the vibration signals generated by laser-induced acoustics into multiple sub-bands, and a set of statistical features is extracted from each frequency band. The GBDT is then utilized to classify and identify these features. The proposed method is validated on a self-built dataset of concrete test blocks with defects. Experimental results demonstrate that the GBDT achieves an accuracy of 90% and a recall rate of 94%. This study confirms the effectiveness of wavelet feature extraction combined with GBDT in laser-induced acoustic detection of concrete defects.
PanYuan, ZhouShuangxi, YangDan. Research on concrete crack segmentation based on improved unet model[J]. Journal of East China Jiaotong University, 2024, 41(1): 11-19. (in Chinese)
TuBing, QinChengjin. Research on debonding defect detection of concrete-filled steel tubular based on SAFT[J]. Journal of Guangxi University (Natural Science Edition), 2025, 50(4): 723-735. (in Chinese)
[5]
ZhengX, LiuF, LuoT, et al. Study on durability and pore characteristics of concrete under salt freezing environment[J]. Materials, 2021, 14(23): 7228.
[6]
RehmanS K U, IbrahimZ, Ali MemonS, et al. Nondestructive test methods for concrete bridges: a review[J]. Construction and Building Materials, 2016, 107: 58-86.
XueYingjie, ChenZhigang, YangRui, et al. Numerical simulation of elliptical cross⁃section projectile penetrating into multi⁃layer concrete target[J]. Journal of North University of China(Natural Science Edition), 2025, 46(1): 35-44. (in Chinese)
ZouJinyang, ZhangYating, DingXin, et al. Detection of internal cavities in concrete with laser-acoustic method[J]. Infrared and Laser Engineering, 2023, 52(1): 134-143. (in Chinese)
[11]
WangK, HeM, FanZ, et al. Impact wave propagation mechanisms and response characteristics in defective concrete[J]. Measurement Science and Technology, 2025, 36(8): 085005.
ZhangRuijie. Concrete image defect detection method based on light weight neural network[D]. Beijing: Beijing University of Chemical Technology, 2025: 9. (in Chinese)
LiHaoxuan, SunYuxiang. Steel surface defect detection based on improved YOLOv8[J]. Journal of Test and Measurement Technology, 2025, 40(1): 1-9. (in Chinese)
ChaiYanli, ZhuYan, WangJianqiang. Wavelet analysis algorithm integrating cyclic autocorrelation function for micro crack detection[J]. Journal of Test and Measurement Technology, 2019, 33(2): 116-120. (in Chinese)
WangJiamin, WuWenhong, NiuHengmao, et al. Review of concrete defect detection methods based on deep learning[J]. Computer Science, 2025, 52(S1): 347-358. (in Chinese)
YuanHong. Numerical simulation study on mechanical effects caused by laser ablation of solid materials[D]. Mianyang: China Academy of Engineering Physics, 2011: 27. (in Chinese)
LiYanlong, WangJunzhong, ChenJunhao, et al. Experimental study on the detection of internal damage of concrete by impact-echo method[J]. Chinese Journal of Applied Mechanics, 2020, 37(1): 149-154. (in Chinese)
[26]
ChoiS, JhangK Y. Internal defect detection using laser-generated longitudinal waves in ablation regime[J]. Journal of Mechanical Science and Technology, 2018, 32(9): 4191-4200.
[27]
KatakuraK, AkamatsuR, SugimotoT, et al. Study on detectable size and depth of defects in noncontact acoustic inspection method[J]. Japanese Journal of Applied Physics, 2014, 53(7S): 07KC15.
[28]
PenedoS R M, NettoM L, JustoJ F. Designing digital filter banks using wavelets[J]. EURASIP Journal on Advances in Signal Processing, 2019(1): 33.
[29]
YanR, GaoR X, ChenX. Wavelets for fault diagnosis of rotary machines: a review with applications[J]. Signal Processing, 2014, 96: 1-15.
[30]
SunB, LiuX. Bearing early fault detection and degradation tracking based on support tensor data description with feature tensor[J]. Applied Acoustics, 2022, 188: 108530.