首页|基于图像融合的碳纤维复合材料冲击损伤检测

基于图像融合的碳纤维复合材料冲击损伤检测

扫码查看
碳纤维复合材料在生产使用过程中,损伤检测一直是关注的一大焦点.锁相热成像检测和超声波C扫描检测是两种常用的检测方式,但各自存在着局限性.超声波C扫描适用于材料内部深层损伤的检测而对浅层损伤信息不敏感,锁相热成像适用于浅层损伤信息的检测,难以反映材料深层次的损伤.针对此问题,该文提出一种基于非下采样剪切波变换结合脉冲耦合神经网络图像融合的算法,将锁相热像图和超声C扫描图像进行图像融合.结果表明,融合图像可以显著提高图像的清晰度和感知质量,并且将损伤面积的误差从39.1%降低到 23.8%.
Impact Damage Detection of Carbon Fiber Reinforced Plastic Based on Image Fusion
Damage detection has been a major focus of attention during the production and use of carbon fiber rein-forced plastic.Lock-in thermography and ultrasonic C-scan are two commonly used detection methods,but each has its own limitations.Ultrasonic C-scan is suitable for the detection of deep damage within the material and is not sen-sitive to shallow damage information,while lock-in thermography is suitable for the detection of shallow damage in-formation,which is difficult to reflect the deep damage of the material.To address this problem,this paper proposes an image fusion algorithm based on non-subsampled shear wave transform combined with pulse-coupled neural net-work to fuse lock-in thermography and ultrasonic C-scan images.The results show that the fused image can signifi-cantly improve the clarity and perceptual quality of the image,and reduce the error of damage area from 39.1%to 23.8%.

lock-in thermal imagingultrasonic C-scanimage fusionnon-subsampled shear wave transform(NSST)car-bon fiber reinforced plastic(CFRP)pulse coupled neural network(PCNN)

张晓龙、程晓颖

展开 >

浙江理工大学机械工程学院,杭州 310018

锁相热成像 超声波C扫描 图像融合 非下采样剪切波变换 碳纤维复合材料 脉冲耦合神经网络

2024

自动化与仪表
天津市工业自动化仪表研究所 天津市自动化学会

自动化与仪表

CSTPCD
影响因子:0.548
ISSN:1001-9944
年,卷(期):2024.39(10)
  • 5