Autonomous Crack Detection using Deep Learning on Synthetic Thermogram Datasets

Chinmay Makarand Pimpalkhare; D. N. Pawaskar

REVIEW 5 major objections 7 minor 20 references

Autonomous Crack Detection using Deep Learning on Synthetic Thermogram Datasets

T0 review · 5 major / 7 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper argues that a synthetic training set of finite-element heat-conduction images of cracked steel plates, rendered as color maps, is enough to train a crack detector after fine-tuning, and that the detector transfers to…

desk verdict A solid in-domain synthetic-data proof-of-concept whose abstract overclaims real-data validation; the wild-image check likely isn't thermographic. read the letter →

arxiv 2412.16499 v1 pith:DVX4J3S5 submitted 2024-12-21 cs.CV

Chinmay Makarand Pimpalkhare , D. N. Pawaskar This is my paper · ORCID

classification cs.CV

keywords crackdetectionthermographysyntheticdatafiniteelementsimulationdomainadaptationdeeplearningimagesegmentationsteelplates

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to show that you do not need thousands of real thermal images to train a deep-learning crack detector for steel plates. It builds a finite-element simulation pipeline that generates steady-state temperature images of plates with randomized cracks and boundary conditions, renders them as color-mapped thermograms, and uses roughly a hundred such images to fine-tune pre-trained vision models. The authors report detection in all 15/15 synthetic test images, with precision 0.996 and recall 0.95 on an earlier split, and they show that the model does transfer to real-world thermograms, but only when crack shape, size, texture, thermal contours, and foreground-background contrast resemble the training data. The practical stake is that expensive and slow thermal experiments could be replaced by fast simulation for building non-destructive-testing datasets.

What carries the argument

The load-bearing mechanism is the randomized finite-element data-generation pipeline: it solves the steady-state heat equation on a rectangular steel plate, represents a crack as an empty void with randomized length, width, location, and inclination, selects boundary conditions (constant temperature or constant heat flux) on each plate edge, and renders the resulting temperature field through one of several color maps so the image resembles a thermogram. The pipeline feeds annotated synthetic images into a pre-trained Mask R-CNN and YOLO variants that are fine-tuned on the small dataset; convolutional edge-detection filters pick up the abrupt local perturbation of thermal contours at the crack, which is the visual signature the model learns. The finite-element solver is benchmarked against an analytical circular-hole solution to validate the heat-conduction model generally, though not for crack-specific thermal signatures.

What would settle it

Take a detector trained only on the synthetic finite-element thermograms and test it on a set of real thermograms of steel plates with known cracks that match the training distribution in crack shape, size, texture, and thermal contours; if detection performance falls to chance on those matched images, the claimed simulation-to-experiment transfer is false.

Watch

Extended reading notes

Core claim

The central claim is that a crack detector can be trained almost entirely on synthetic thermograms generated by solving the steady-state heat equation on a steel plate in which a crack is modeled as a region of removed material. After fine-tuning a pre-trained instance-segmentation model on 105 synthetic images (with 30 for validation and 15 for testing), the model detected the crack in all 15 test images, including cases a human would struggle to see; an earlier object-detection experiment reported precision 0.996, recall 0.95, and mAP50 0.947 on an 80/20 split. The paper also claims that translation to experimental 'wild' images succeeds under matching conditions—similar crack shape, size, image texture, thermal contours, and foreground-background differences—and fails when those are absent, for example with ordinary grayscale images. This establishes the conditions under which simulation-based training data can substitute for experimental thermograms.

Load-bearing premise

The whole approach depends on simulated temperature images looking enough like real thermograms that a model trained on them keeps working on experimental data—a premise the paper itself flags as fragile because simulated images are 'very clean and crisp' compared with real ones.

Editorial extensions

If this is right

A synthetic training set can be generated in minutes (about 100 images in 3 minutes), removing the thermal-experiment bottleneck for building crack-detection datasets.
Fine-tuning a pre-trained model on about a hundred synthetic images is enough to detect cracks in all synthetic test images and to transfer to real thermograms under matching appearance conditions.
Transfer is conditional rather than automatic: real images must resemble the synthetic training distribution in crack shape, size, texture, thermal contours, and foreground-background contrast, and occluded cracks will be missed.
Adding noise to the clean synthetic images, as the paper suggests, is a direct way to widen the range of real thermograms the model can handle.
Because the detector does not generalize to ordinary grayscale crack images, the learned representation is specific to the thermal modality rather than to cracks in general.
A direct experimental measurement of the thermal perturbation around a known crack would test whether the material-removal crack model is physically faithful enough for training data; the paper validates the solver against a circular-hole solution, not against a crack-specific signature.
A controlled study that adds measured sensor noise, blur, and emissivity variation to synthetic renderings could isolate which rendering choices control the sim-to-real gap and turn the paper's qualitative wild-image observations into a quantitative rule.
The same pipeline could be extended to transient heating, three-dimensional geometries, subsurface cracks, or cracks modeled as regions of altered material properties; the paper lists these as future work but does not test them.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Pimpalkhare and Pawaskar report a MATLAB FEM pipeline that generates synthetic thermogram-like images of cracked steel plates, then fine-tune Detectron2/YOLO on ~100 images and get high in-domain detection numbers (precision 0.996, recall 0.95, mAP50 0.947; 15/15 on the held-out synthetic set). That part is believable and reasonably well described. The pipeline parameters (crack length, width, angle, boundary conditions, colormaps) are randomized sensibly, and the authors are candid about limitations in IX.B: they note the synthetic images are 'clean and crisp' and that texture/granularity mismatch hurts transfer.

The soft spot is the abstract's claim that the approach 'translates to realistic experimental data.' The only evidence is the 'domain adaptation check' on 'wild' images, and the acquisition modality of those images is never stated. The model failed on grayscale images even though the training set included grayscale colormaps; that strongly suggests the wild set was visible-light photos, not thermograms. If so, the transfer to real thermal NDT is entirely unvalidated. The FEM benchmark in IV.B is against a circular-hole analytical solution, not a crack, so the crack-specific thermal signature is unverified. Add to that: metrics from single runs, no error bars, no code/data release, and a dataset of only 80-105 training images. For an in-domain claim those are minor; for the transfer claim they are load-bearing.

I think the paper is honest about its own limits in the body, but the abstract and the word 'validated' go beyond what IX.B can support. The novelty is moderate—Kovacs et al. and Fang et al. already did synthetic thermal simulations for defect detection—but the steel-plate FEM variant and the qualitative failure-mode list are a legitimate extension.

Who gets value? Someone building synthetic NDT pipelines might read it to avoid the same domain-adaptation pitfalls. It deserved a serious referee rather than desk rejection, because the core idea is sound and the limitations are fixable with one modest real-thermogram experiment and a clearer statement about the wild images. My recommendation: ask for a revision that either obtains a small set of real thermograms or explicitly retracts the 'realistic experimental data' claim.

Referee Report

5 major / 7 minor

Summary. The manuscript presents a synthetic data generation pipeline for crack detection in steel plates, based on steady-state finite element simulations of thermal profiles with cracks modeled as material removal. The generated temperature maps are rendered with MATLAB colormaps (jet, inferno, grayscale) and augmented. The authors fine-tune pre-trained object detection and segmentation models (YOLO, Detectron2/Mask R-CNN) on 80–105 synthetic images and report high precision/recall and mAP50 on an in-domain synthetic test set (15/15 correct detections). They also report a qualitative "domain adaptation" check on images from the wild, concluding that the approach can translate to realistic experimental data.

Significance. If fully supported, the claim would be significant: it would show that a very small synthetic dataset can replace expensive and difficult thermographic experiments for training crack detectors. The in-domain metrics (Precision 0.996, Recall 0.95, mAP50 0.947; Section VI.B) and the 15/15 detection rate on synthetic test data (Section IX.A) are promising and give a reproducible baseline, especially since Algorithm 1 provides a concrete pipeline and the reported generation speed (100 images in 3 minutes, Section III.A) makes the in-domain setup easy to replicate. However, the paper's headline claim of translation to realistic experimental data rests entirely on a qualitative check on about ten images of unspecified acquisition modality (Section IX.B), and the paper's own observations contradict the suitability of that check. The FEM benchmark (Section IV.B) validates the solver for a circular hole, not for the crack geometry used in training. The external-validity claim is therefore currently unsupported.

major comments (5)

[Abstract and Section IX.B] The abstract states that the authors 'validated the results by checking if our approach translates to realistic experimental data,' but Section IX.B provides no evidence that the 'wild' images are thermograms. The section never names the acquisition modality; it only says they were 'picked up from real-world scenarios.' Observation (d) says the model 'could not scale onto crack detection scenarios where the images are simply grayscale,' even though Section III.B says the synthetic training set included grayscale colormaps. This mismatch strongly suggests the wild set consisted of visible-light grayscale images, not experimental thermograms. The claimed validation on realistic experimental thermographic data is therefore not established.
[Section IX.B and Table I] The domain-adaptation evaluation is entirely qualitative: it reports performance on 'of the order of 10^1' images, with no exact count, no quantitative metrics (e.g., mAP, IoU), and no description of the decision procedure behind the ✓/✗ entries in Table I. A qualitative check on roughly ten images cannot support the abstract's generalization claim, especially when the model is reported to fail on grayscale inputs that were part of the training distribution.
[Section IV.B] The benchmarking against the analytical solution for a circular hole validates the FEM solver's accuracy for a hole, not for the crack geometry (defined as material removal with varying width, length, and inclination) used to generate the training data. The crack-specific thermal signature is the very feature the detector must learn; without a crack-specific validation, the simulation's fidelity for the target defect is unverified. At minimum, a convergence study on the crack geometry or a comparison with a known crack solution is needed.
[Sections VI.A, VI.B, and IX.A] All quantitative detection results are obtained on synthetic test images drawn from the same generation pipeline as the training data. The 15/15 result in Section IX.A is on this in-domain test set, and the metrics in Section VI.B (Precision 0.996, Recall 0.95, mAP50 0.947) are for the same condition. These results support the weak claim that a detector can be trained on synthetic thermograms, but they do not support the paper's broader transfer claim; the manuscript should clearly separate in-domain performance from any cross-domain evidence.
[Section XI] The future-work section concedes that 'How our created images directly map to the images obtained through experimental images is also an avenue to check in the future.' This is an explicit admission that the mapping from synthetic to real thermograms has not been established. Given that the abstract claims such a validation, the manuscript is internally inconsistent about what has been demonstrated.

minor comments (7)

[Abstract] The abstract contains a typo: 'Convolutional Neural Netowrks' should be 'Networks', and the sentence beginning 'There has been a rise in the use of Artificial Intelligence...' is grammatically incomplete.
[Section III.B] 'CV AT.io' should be 'CVAT.io'.
[Section IV] The 'Characteristic Range' table is presented as 'Fig. 8' but is a table; it should be numbered as a table with a caption.
[Section V.C] The text mentions 'RCNN models,' then 'YOLO series of models,' then 'Detectron2' without clarifying which architectures were actually fine-tuned for the reported experimental results.
[Sections VI.A and VIII.B] Section VI.A reports an 80-image training dataset, while Section VIII.B states a 105-image training dataset; the relationship between these configurations and the metrics reported in Section VI.B is not explained.
[Section IX.B] The phrase 'the number of images that we could procure were of the order of 10^1' should use 'was' instead of 'were,' and the exact number of wild images should be stated.
[Figures] Figure numbering is inconsistent: Fig. 8 is a table, and Figs. 5 and 6 (easy/hard examples) appear in the text without clear in-order call-outs.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the synthetic-to-real transfer claim is under-validated but not constructed from its own outputs.

full rationale

The paper's derivation chain is self-contained. Synthetic thermogram-like images are generated from finite element simulations (Section III.B), rendered with MATLAB colormaps, annotated, and used to fine-tune pretrained detectors (Sections V and VIII). The held-out synthetic test set in Section IX.A measures in-distribution generalization; no fitted parameter is renamed as a prediction. The analytical circular-hole benchmark in Section IV.B independently validates the FEM solver for a hole geometry, and although it does not validate the crack-specific thermal signature, this is a soundness limitation rather than circularity. No load-bearing step relies on the authors' own prior work; all cited deep-learning and synthetic-data references are external. The main weaknesses—the undisclosed acquisition modality of the 'wild' images in Section IX.B, the questionable transfer from synthetic thermograms to real thermographic NDT, and the observed failure on grayscale images despite grayscale training samples—are external-validity and generalization concerns, not cases where the result reduces to its own inputs by definition or by self-citation. Therefore no circular step can be exhibited, and the circularity score is 0.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on domain assumptions about FEM fidelity, crack modeling, and transferability of pre-trained features, plus hand-chosen data-generation ranges. No invented physical entities are introduced.

free parameters (5)

Crack width exponent range = 10^-2 to 10^-4 units
Chosen by hand in Section IV to mix easy (thick) and hard (thin) crack examples; no relation to real crack statistics.
Crack length sampling range = [0.3, 0.7] units
Uniform sampling range chosen in Algorithm 1; not derived from measurements.
Crack inclination sampling range = [0, 2*pi]
Uniform sampling range; design choice to maximize diversity.
Boundary condition types = constant temperature or constant heat flux per edge
Randomized per edge in Algorithm 1; no experimental grounding for which condition matches real thermography setups.
Colormap set = jet, inferno, grayscale
Selected in Section III.B because they looked 'realistic'; hand-chosen visual prior.

assumptions (4)

domain assumption Steady-state FEM heat conduction is an accurate model of thermographic temperature fields.
Invoked implicitly in Section III.B. The solver is benchmarked against an analytical solution only for a circular hole, not for cracks.
domain assumption A crack can be modeled as a region of removed material (empty space).
Stated in Section III.B: 'We define a crack as a region where there is a loss in the base material'. Real cracks have complex geometry and partial closure.
domain assumption Analytical validation on a circular-hole geometry transfers to crack geometries.
Section IV.B benchmarks only the circular hole configuration; the extension to cracks is assumed without crack-specific analytical or experimental validation.
domain assumption Pre-trained natural-image features transfer to synthetic thermograms.
Sections V.C and VII.A assume that a backbone pre-trained on natural images, after fine-tuning, captures thermogram crack features.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Autonomous Crack Detection using Deep Learning on Synthetic Thermogram Datasets." pith.science (2026). https://pith.science/paper/DVX4J3S5

@misc{pith2026241216499,
  author       = {Pith},
  title        = {Pith review of: Autonomous Crack Detection using Deep Learning on Synthetic Thermogram Datasets},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DVX4J3S5}},
  note         = {Machine review of arXiv:2412.16499}
}

read the original abstract

In a lot of scientific problems, there is the need to generate data through the running of an extensive number of experiments. Further, some tasks require constant human intervention. We consider the problem of crack detection in steel plates. The way in which this generally happens is through humans looking at an image of the thermogram generated by heating the plate and classifying whether it is cracked or not. There has been a rise in the use of Artificial Intelligence (AI) based methods which try to remove the requirement of a human from this loop by using algorithms such as Convolutional Neural Netowrks (CNN)s as a proxy for the detection process. The issue is that CNNs and other vision models are generally very data-hungry and require huge amounts of data before they can start performing well. This data generation process is not very easy and requires innovation in terms of mechanical and electronic design of the experimental setup. It further requires massive amount of time and energy, which is difficult in resource-constrained scenarios. We try to solve exactly this problem, by creating a synthetic data generation pipeline based on Finite Element Simulations. We employ data augmentation techniques on this data to further increase the volume and diversity of data generated. The working of this concept is shown via performing inference on fine-tuned vision models and we have also validated the results by checking if our approach translates to realistic experimental data. We show the conditions where this translation is successful and how we can go about achieving that.

Figures

Figures reproduced from arXiv: 2412.16499 by the authors.

**Figure 1.** Overview of our solution Our approach involves three important aspects. The first is to create a data generation pipeline utilizing finite element simulations. The code has been written in MATLAB and has been automated such that it is possible to generate data quickly. At the moment, we are able to generate 100 images in 3 minutes, which is fast enough for most application since this data can be created and stored i… view at source ↗

**Figure 2.** The Data Generation Pipeline In this section, we shall breifly explain the data generation pipeline and also discuss why specific design choices were made. The data which we are using is being generated through finite element simulations. The finite element method (FEM) is a popular method for numerically solving differential equations arising in engineering and mathematical modeling. It is a generally used for solv… view at source ↗

**Figure 3.** Simulation Template The first design choice is regarding the use of the base geometry. Since most of the objects we would study were in the form of plates, we decided to choose a rectangular outer boundary for the object. The outer boundary is not very important, and a circular one could be used too. However, using a polygonal boundary allows us to effectively define a single boundary condition at every side of the … view at source ↗

Figures from the paper (8 more)

**Figure 4.** Figure 4: From left to right −→ (a) Original image. (b) Image with high Brightness. (c) Image with parts near the crack blurred to decrease visibility. (d) Image with contrast, brightness and exposure adjusted such that it is harder to locate the crack [PITH_FULL_IMAGE:figures/…

**Figure 7.** Figure 7: Improved Simulation Template The earlier pipeline was creating a lot of redundant images because there were only horizontal cracks. We wish to improve the data generation by applying a greater variety of boundary conditions. We also increase the diversity by randomly s…

**Figure 6.** Figure 6: Hard Examples Easier examples are characterized by conspicuous variations in the thermal profile due to the presence of a crack. These instances are readily discernible, making it relatively straightforward for both automated systems and the human eye to identify the …

**Figure 8.** Figure 8: Examples generated by new pipeline A. Deep Learning Pipeline [PITH_FULL_IMAGE:figures/full_fig_p005_8.png]

**Figure 10.** Figure 10: Benchmarking A thorough comparison for a circular hole geometry was conducted, revealing a high level of agreement between the analytical and numerical solutions. The visual examination indicates a close match between the two, with magnitudes exhibiting a striking sim…

**Figure 11.** Figure 11: Local Features from Neural Networks B. Edge Detection based on Convolution This is a key operation happening in the background of convolutional networks which is extremely important in our case. This is because cracks are generally similar to onedimensional objects a…

**Figure 12.** Figure 12: Edge Detection Kernel The edge detection kernel applied to the input image generates an output feature map that predominantly appears black, with the exception of the area corresponding to the location of the crack. C. Architecture We implement RCNN models for the ta…

**Figure 14.** Figure 14: Inference Results The inference results are very promising. Here, we just grade the model on basis of whether it was able to detect the crack or not. We observe that in each of the 15/15 cases, the model was able to detect the crack correctly. This is excellent as the…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

20 extracted references · 18 canonical work pages

[1]

Jaeger, B.E., Schmid, S., Grosse, C.U. et al. Infrared Thermal Imaging- Based Turbine Blade Crack Classification Using Deep Learning. J Nondestruct Eval 41, 74 (2022). https://doi.org/10.1007/s10921-022- 00907-9

work page doi:10.1007/s10921-022- 2022
[2]

Yang, J., Wang, W., Lin, G., Li, Q., Sun, Y ., & Sun, Y . (2019). Infrared Thermal Imaging-Based Crack Detection Using Deep Learning. IEEE Access, 7, 182060-182077

work page 2019
[3]

Arun Mohan, Sumathi Poobal, Crack detection using image process- ing: A critical review and analysis, Alexandria Engineering Jour- nal, V olume 57, Issue 2, 2018, Pages 787-798, ISSN 1110-0168, https://doi.org/10.1016/j.aej.2017.01.020

work page doi:10.1016/j.aej.2017.01.020 2018
[4]

Tian, L., Wang, Z., Liu, W. et al. A New GAN-Based Approach to Data Augmentation and Image Segmentation for Crack Detection in Thermal Imaging Tests. Cogn Comput 13, 1263–1273 (2021). https://doi.org/10.1007/s12559-021-09922-w

work page doi:10.1007/s12559-021-09922-w 2021
[5]

Alexander, Q.G., Hoskere, V ., Narazaki, Y . et al. Fusion of thermal and RGB images for automated deep learning based crack detection in civil infrastructure. AI Civ. Eng. 1, 3 (2022). https://doi.org/10.1007/s43503- 022-00002-y

work page doi:10.1007/s43503- 2022
[6]

Deep Learning Based Infrared Thermal Image Analysis of Complex Pavement Defect Conditions Considering Seasonal Effect

Chandra S, AlMansoor K, Chen C, Shi Y , Seo H. Deep Learning Based Infrared Thermal Image Analysis of Complex Pavement Defect Conditions Considering Seasonal Effect. Sensors. 2022; 22(23):9365. https://doi.org/10.3390/s22239365

work page doi:10.3390/s22239365 2022
[7]

P ´eter Kov ´acs, Bernhard Lehner, Gregor Thummerer, G ¨unther Mayr, Peter Burgholzer, Mario Huemer; Deep learning approaches for ther- mographic imaging. J. Appl. Phys. 21 October 2020; 128 (15): 155103. https://doi.org/10.1063/5.0020404

work page doi:10.1063/5.0020404 2020
[8]

Automatic Defects Segmen- tation and Identification by Deep Learning Algorithm with Pulsed Ther- mography: Synthetic and Experimental Data

Fang Q, Ibarra-Castanedo C, Maldague X. Automatic Defects Segmen- tation and Identification by Deep Learning Algorithm with Pulsed Ther- mography: Synthetic and Experimental Data. Big Data and Cognitive Computing. 2021; 5(1):9. https://doi.org/10.3390/bdcc5010009

work page doi:10.3390/bdcc5010009 2021

Show all 20 references

[9]

Matlab Official Documentation: https://in.mathworks.com/help/matlab/
[10]

YOLOv5 Official Implementation: https://github.com/ultralytics/yolov5
[11]

J., & Shotton, J

Wood, E., Baltru ˇsaitis, T., Hewitt, C., Dziadzio, S., Cashman, T. J., & Shotton, J. (2021). Fake it till you make it: face analysis in the wild using synthetic data alone. Proceedings of the IEEE/CVF International Conference on Computer Vision, 3681–3691

2021
[12]

He, R., Sun, S., Yu, X., Xue, C., Zhang, W., Torr, P., Bai, S., & Qi, X. (2022). Is synthetic data from generative models ready for image recognition? ArXiv Preprint arXiv:2210.07574

2022 arXiv
[13]

Nikolenko, S. I. (2021). Synthetic data for deep learning (V ol. 174). Springer

2021
[14]

Wang, Q., Gao, J., Lin, W., & Yuan, Y . (2021). Pixel-wise crowd understanding via synthetic data. International Journal of Computer Vision, 129(1), 225–245

2021
[15]

Hahner, M., Sakaridis, C., Dai, D., & Van Gool, L. (2021). Fog Simula- tion on Real LiDAR Point Clouds for 3D Object Detection in Adverse Weather. Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 15283–15292

2021
[16]

Lu, Y ., Shen, M., Wang, H., Wang, X., van Rechem, C., Fu, T., & Wei, W. (2023). Machine learning for synthetic data generation: a review. ArXiv Preprint arXiv:2302.04062

2023 arXiv
[17]

A., Paczan, N., Webb, R., & Susskind, J

Roberts, M., Ramapuram, J., Ranjan, A., Kumar, A., Bautista, M. A., Paczan, N., Webb, R., & Susskind, J. M. (2021). Hypersim: A Pho- torealistic Synthetic Dataset for Holistic Indoor Scene Understanding. Proceedings of the IEEE/CVF International Conference on Computer Vision (...

2021
[18]

Wu, X., Hong, D., & Chanussot, J. (2022). UIU-Net: U-Net in U-Net for infrared small object detection. IEEE Transactions on Image Processing, 32, 364–376

2022
[19]

Tokmakov, P., Li, J., Burgard, W., & Gaidon, A. (2021). Learning To Track With Object Permanence. Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 10860–10869

2021
[20]

Jain, S., Seth, G., Paruthi, A., Soni, U., & Kumar, G. (2022). Synthetic data augmentation for surface defect detection and classification using deep learning. Journal of Intelligent Manufacturing, 1–14

2022

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

Jaeger, B.E., Schmid, S., Grosse, C.U. et al. Infrared Thermal Imaging- Based Turbine Blade Crack Classification Using Deep Learning. J Nondestruct Eval 41, 74 (2022). https://doi.org/10.1007/s10921-022- 00907-9

work page doi:10.1007/s10921-022- 2022

[2] [2]

Yang, J., Wang, W., Lin, G., Li, Q., Sun, Y ., & Sun, Y . (2019). Infrared Thermal Imaging-Based Crack Detection Using Deep Learning. IEEE Access, 7, 182060-182077

work page 2019

[3] [3]

Arun Mohan, Sumathi Poobal, Crack detection using image process- ing: A critical review and analysis, Alexandria Engineering Jour- nal, V olume 57, Issue 2, 2018, Pages 787-798, ISSN 1110-0168, https://doi.org/10.1016/j.aej.2017.01.020

work page doi:10.1016/j.aej.2017.01.020 2018

[4] [4]

Tian, L., Wang, Z., Liu, W. et al. A New GAN-Based Approach to Data Augmentation and Image Segmentation for Crack Detection in Thermal Imaging Tests. Cogn Comput 13, 1263–1273 (2021). https://doi.org/10.1007/s12559-021-09922-w

work page doi:10.1007/s12559-021-09922-w 2021

[5] [5]

Alexander, Q.G., Hoskere, V ., Narazaki, Y . et al. Fusion of thermal and RGB images for automated deep learning based crack detection in civil infrastructure. AI Civ. Eng. 1, 3 (2022). https://doi.org/10.1007/s43503- 022-00002-y

work page doi:10.1007/s43503- 2022

[6] [6]

Deep Learning Based Infrared Thermal Image Analysis of Complex Pavement Defect Conditions Considering Seasonal Effect

Chandra S, AlMansoor K, Chen C, Shi Y , Seo H. Deep Learning Based Infrared Thermal Image Analysis of Complex Pavement Defect Conditions Considering Seasonal Effect. Sensors. 2022; 22(23):9365. https://doi.org/10.3390/s22239365

work page doi:10.3390/s22239365 2022

[7] [7]

P ´eter Kov ´acs, Bernhard Lehner, Gregor Thummerer, G ¨unther Mayr, Peter Burgholzer, Mario Huemer; Deep learning approaches for ther- mographic imaging. J. Appl. Phys. 21 October 2020; 128 (15): 155103. https://doi.org/10.1063/5.0020404

work page doi:10.1063/5.0020404 2020

[8] [8]

Automatic Defects Segmen- tation and Identification by Deep Learning Algorithm with Pulsed Ther- mography: Synthetic and Experimental Data

Fang Q, Ibarra-Castanedo C, Maldague X. Automatic Defects Segmen- tation and Identification by Deep Learning Algorithm with Pulsed Ther- mography: Synthetic and Experimental Data. Big Data and Cognitive Computing. 2021; 5(1):9. https://doi.org/10.3390/bdcc5010009

work page doi:10.3390/bdcc5010009 2021