Image inpainting in acoustic microscopy
Source abstract
Abstract
Scanning acoustic microscopy (SAM) is a non-ionizing and label-free imaging modality used to visualize the surface and internal structures of industrial objects and biological specimens. The image of the sample under investigation is created using high-frequency acoustic waves. The frequency of the excitation signals, the signal-to-noise ratio, and the pixel size all play a role in acoustic image resolution. We propose a deep learning-enabled image inpainting for acoustic microscopy in this paper. The method is based on training various generative adversarial networks (GANs) to inpaint holes in the original image and generate a 4× image from it. In this approach, five different types of GAN models are used: AOTGAN, DeepFillv2, Edge-Connect, DMFN, and Hypergraphs image inpainting. The trained model’s performance is assessed by calculating the peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) between network-predicted and ground truth images. The Hypergraphs image inpainting model provided an average SSIM of 0.93 for 2× and up to 0.93 for the final 4×, respectively, and a PSNR of 32.33 for 2× and up to 32.20 for the final 4×. The developed SAM and GAN frameworks can be used in a variety of industrial applications, including bio-imaging.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Image and Signal Denoising Methods · Advanced Image Processing Techniques · Generative Adversarial Networks and Image Synthesis
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Department of Mathematics, Indian Institute of Technology Guwahati 1 , Guwahati 781039 Assam, India
Department of Physics, Indian Institute of Technology Delhi 2 , Delhi, India
Department of Physics and Technology, UiT The Arctic University of Norway 3 , 9037 Tromsø, Norway
Department of Computer Science, UiT The Arctic University of Norway 4 , 9037 Tromsø, Norway
Funding
Research Council of Norway · Cristin Project
References (52)
Show source-provided references
1. 2010
2. Scanning acoustic microscopy—A novel noninvasive method to determine tumor interstitial fluid pressure in a xenograft tumor model10.1016/j.tranon.2016.03.009 ↗
3. Quantitative ultrasonic characterization of c-axis oriented polycrystalline AlN thin film for smart device application10.3813/aaa.918863 ↗
4. Chirp coded ultrasonic pulses used for scanning acoustic microscopy
5. Mechanical characterization of sintered piezo-electric ceramic material using scanning acoustic microscope10.1016/j.ultras.2012.07.011 ↗
6. Ultrasonic measurements of surface defects on flexible circuits using high-frequency focused polymer transducers10.7567/jjap.56.07jc05 ↗
7. Non-destructive evaluation (NDE) of aerospace composites: Acoustic microscopy10.1533/9780857093554.3.423 ↗
8. In vitro volume imaging of articular cartilage using chirp-coded high frequency ultrasound
9. Improved failure analysis in scanning acoustic microscopy via advanced signal processing techniques10.1016/j.microrel.2022.114618 ↗
10. Inspection of multilayered electronic devices via scanning acoustic microscopy using synthetic aperture focusing technique
11. 1993
12. Image inpainting
13. Image inpainting based on deep learning: A review10.1016/j.displa.2021.102028 ↗
14. Generative image inpainting with contextual attention
15. Score-based generative modeling through stochastic differential equations
16. Image inpainting for irregular holes using partial convolutions
17. Deep image prior
18. Region filling and object removal by exemplar-based image inpainting10.1109/tip.2004.833105 ↗
19. Fragment-based image completion10.1145/882262.882267 ↗
20. PatchMatch: A randomized correspondence algorithm for structural image editing10.1145/1531326.1531330 ↗
21. Texture synthesis by non-parametric sampling
22. Photo-realistic single image super-resolution using a generative adversarial network
23. Edge-informed single image super-resolution10.1109/ICCVW.2019.00409 ↗
24. Context encoders: Feature learning by inpainting10.1109/CVPR.2016.278 ↗
25. Free-form image inpainting with gated convolution
26. Edgeconnect: Generative image inpainting with adversarial edge learning
27. Deep semantic face deblurring
28. Deep colorization10.1109/ICCV.2015.55 ↗
29. Image inpainting with local and global refinement10.1109/tip.2022.3152624 ↗
30. Deep learning for image inpainting: A survey10.1016/j.patcog.2022.109046 ↗
31. Image inpainting based on generative adversarial networks10.1109/access.2020.2970169 ↗
32. U. Demir and G.Unal, “Patch-based image inpainting with generative adversarial networks,” arXiv:1803.07422 (2018).
33. Image multi-inpainting via progressive generative adversarial networks
34. Globally and locally consistent image completion10.1145/3072959.3073659 ↗
35. Shift-net: Image inpainting via deep feature rearrangement
36. Hyperrealistic image inpainting with hypergraphs
37. Aggregated contextual transformations for high-resolution image inpainting10.1109/TVCG.2022.3156949 ↗
38. Image fine-grained inpainting
39. Hyperrealistic image inpainting with hypergraphs
40. Progressive growing of GANs for improved quality, stability, and variation
41. 2007 · 1601
42. Hypergraph neural networks10.1609/aaai.v33i01.33013558 ↗
43. HyperGCN: A new method of training graph convolutional networks on hypergraphs
44. Squeeze-and-excitation networks10.1109/CVPR.2018.00745 ↗
45. Fast and accurate deep network learning by exponential linear units (ELUs)
46. Image-to-image translation with conditional adversarial networks
47. MSCE: An edge preserving robust loss function for improving super-resolution algorithms10.1007/978-3-030-04224-0_49 ↗
48. Tensorflow, High-quality version of the CelebA dataset, consisting of 30 000 images in 1024 × 1024 resolution, https://www.tensorflow.org/datasets/catalog/celeb_a_hq; accessed 14 March 2023.
49. Progressive growing of GANs for improved quality, stability, and variation
50. Standa, Motorized xy microscope stage-motorized positioners and controllers-catalog-opto-mechanical products-standa, https://www.standa.lt/products/catalog/motorised_positioners?item=609&prod=motorized_xy_microscope_stage, 2022, accessed 16 March 2022.
51. Numerical method for tilt compensation in scanning acoustic microscopy10.1016/j.measurement.2021.110306 ↗
52. Becker, Becker nachrichtentechnik gmbh (bnt), “1 W wideband Amplifier,” https://www.becker-rf.com/files_db/1608213582_2017_17.pdf, 2022, accessed 16 March 2022.