Generative Principal Component Thermography for Enhanced Defect Detection and Analysis

Liu, Kaixin; Li, Yingjie; Yang, Junxiao; Liu, Yi; Yao, Yuan

doi:10.1109/tim.2020.2992873

Cited by 62 publications

(42 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Meanwhile, the fraction of subgrain boundary is relatively high in BM (Figure 15). The subgrain boundaries were also treated as the barriers against the dislocation movements, improving the strength and fatigue resistance of BM [23,24]. However, the fatigue crack propagation was determined by a transgranular-intergranular mixed mode.…”

Section: Discussionmentioning

confidence: 99%

Fatigue Crack Growth Characteristics of 34CrMo4 Steel for Gas Cylinders by Cold Flow Forming after Hot Drawing

et al. 2021

Metals

View full text Add to dashboard Cite

The fatigue crack growth (FCG) behavior of 34CrMo4 steel, a typical material for gas cylinders, has been investigated. Specimens were taken from the base material (BM) as well as the hot-drawn (HD) cylinder and cold-flow (CF) formed cylinder along the longitudinal and transverse directions. The FCG tests were conducted under different stress ratios for different materials and directions. The main purpose of this research was to explore the influences of the mechanical and thermal processes, sampling direction and stress ratio on the FCG behavior of 34CrMo4 steel. To further reveal the mechanism of crack propagation at different stages, the microstructures and fracture modes of FCG specimens were analyzed by scanning electron microscopy (SEM) and electron backscatter diffraction (EBSD), respectively. The results showed that HD and CF materials exhibited better resistance to fatigue crack propagation than BM. The FCG rates of investigated materials can be accelerated by the increase in stress ratio. However, the sampling direction had little effect on the FCG rate. Finally, a driving force parameter (DFP) model was used to fit the experimental FCG data of three materials with different mechanical and thermal processes. A unified transition stage between the stable and unstable FCG stages of three materials under various experimental conditions was revealed by DFP model, playing an important role on the early warning of fatigue fracture for different types of 34CrMo4 steel.

show abstract

Section: Discussionmentioning

confidence: 99%

Fatigue Crack Growth Characteristics of 34CrMo4 Steel for Gas Cylinders by Cold Flow Forming after Hot Drawing

et al. 2021

Metals

View full text Add to dashboard Cite

show abstract

“…Recently, Liu et al [ 31 ] introduced an approach that uses data augmentation generated by the deep-learning models. The assumption was that deep-learning models would be able to learn statistical features from the data.…”

Section: Literature Reviewmentioning

confidence: 99%

Robust Principal Component Thermography for Defect Detection in Composites

Ebrahimi

Fleuret

Klein

et al. 2021

Sensors

View full text Add to dashboard Cite

Pulsed Thermography (PT) data are usually affected by noise and as such most of the research effort in the last few years has been directed towards the development of advanced signal processing methods to improve defect detection. Among the numerous techniques that have been proposed, principal component thermography (PCT)—based on principal component analysis (PCA)—is one of the most effective in terms of defect contrast enhancement and data compression. However, it is well-known that PCA can be significantly affected in the presence of corrupted data (e.g., noise and outliers). Robust PCA (RPCA) has been recently proposed as an alternative statistical method that handles noisy data more properly by decomposing the input data into a low-rank matrix and a sparse matrix. We propose to process PT data by RPCA instead of PCA in order to improve defect detectability. The performance of the resulting approach, Robust Principal Component Thermography (RPCT)—based on RPCA, was evaluated with respect to PCT—based on PCA, using a CFRP sample containing artificially produced defects. We compared results quantitatively based on two metrics, Contrast-to-Noise Ratio (CNR), for defect detection capabilities, and the Jaccard similarity coefficient, for defect segmentation potential. CNR results were on average 40% higher for RPCT than for PCT, and the Jaccard index was slightly higher for RPCT (0.7395) than for PCT (0.7010). In terms of computational time, however, PCT was 11.5 times faster than RPCT. Further investigations are needed to assess RPCT performance on a wider range of materials and to optimize computational time.

show abstract

“…For IRT defect identification, several thermographic data analysis methods have been developed to enhance the visibility of defects in thermograms. These methods include principal component thermography (PCT) [ 10 ] and its variants (such as sparse PCT [ 11 ], candid covariance-free incremental PCT [ 12 ], generative PCT [ 13 ]), thermographic signal reconstruction [ 14 ], penalized least squares [ 15 ], pulsed phase thermography [ 16 ], independent component thermography [ 17 ], manifold learning methods [ 18 , 19 , 20 ], autoencoder methods [ 21 , 22 ], other deep learning methods [ 23 , 24 , 25 ], etc. Most of the above-mentioned methods can be considered as a kind of metric learning [ 26 ].…”

Section: Introductionmentioning

confidence: 99%

“…Another challenge of thermographic data analysis is that the thermograms are typically limited because of the shallow detection depth of the thermal imager, which indeed restricts the defect detection capabilities of thermal imaging methods. While this difficulty is solved by using a data expansion strategy in the generative PCT (GPCT) approach [ 13 ], there is still room for improvement in the performance of defect identification because the nonlinearity of thermal image data is not considered in it.…”

Section: Introductionmentioning

confidence: 99%

Enhanced Defect Detection in Carbon Fiber Reinforced Polymer Composites via Generative Kernel Principal Component Thermography

Liu

et al. 2021

Polymers

Self Cite

View full text Add to dashboard Cite

Increasing machine learning methods are being applied to infrared non-destructive assessment for internal defects assessment of composite materials. However, most of them extract only linear features, which is not in accord with the nonlinear characteristics of infrared data. Moreover, limited infrared images tend to restrict the data analysis capabilities of machine learning methods. In this work, a novel generative kernel principal component thermography (GKPCT) method is proposed for defect detection of carbon fiber reinforced polymer (CFRP) composites. Specifically, the spectral normalization generative adversarial network is proposed to augment the thermograms for model construction. Sequentially, the KPCT method is used by feature mapping of all thermogram data using kernel principal component analysis, which allows for differentiation of defects and background in the dimensionality-reduced data. Additionally, a defect-background separation metric is designed to help the performance evaluation of data analysis methods. Experimental results on CFRP demonstrate the feasibility and advantages of the proposed GKPCT method.

show abstract

Generative Principal Component Thermography for Enhanced Defect Detection and Analysis

Cited by 62 publications

References 27 publications

Fatigue Crack Growth Characteristics of 34CrMo4 Steel for Gas Cylinders by Cold Flow Forming after Hot Drawing

Fatigue Crack Growth Characteristics of 34CrMo4 Steel for Gas Cylinders by Cold Flow Forming after Hot Drawing

Robust Principal Component Thermography for Defect Detection in Composites

Enhanced Defect Detection in Carbon Fiber Reinforced Polymer Composites via Generative Kernel Principal Component Thermography

Contact Info

Product

Resources

About