In ultrasonic nondestructive testing, the low resolution of ultrasound images possibly lead to misinterpretation of defects in the image. At present, there is no special data set for ultrasonic nondestructive testing images in super-resolution, and the performance of numerous existing models depends on the learning of general data sets. In this paper, a zero-shot super-resolution network based on reflection projection units is proposed. Ultrasound images contain numerous image blocks with similar content, which are randomly extracted and down-sampled to form training samples. Then the image features are extracted through the reflection projection units in the network, and the information between the high and low-resolution image pairs is fully excavated. Finally, the feature channel is reduced by the attention mechanism, and the reconstructed image is output. Moreover, a combined loss function is used to optimize the network parameters. The compared experiments show that the proposed method performs better than the state of the art. Code is available at:https://github.com/BJUT-VIP/Zero-Shot-UNDTSR.
Zero-Shot Ultrasound Nondestructive Testing Image Super-Resolution Based on Reflection Projection
Lect. Notes Electrical Eng.
International Conference on Autonomous Unmanned Systems ; 2021 ; Changsha, China September 24, 2021 - September 26, 2021
Proceedings of 2021 International Conference on Autonomous Unmanned Systems (ICAUS 2021) ; Chapter : 129 ; 1287-1297
2022-03-18
11 pages
Article/Chapter (Book)
Electronic Resource
English
"Zero-Shot" Super-Resolution Based on SinGAN
IEEE | 2021
|Projection registration applied to nondestructive testing
British Library Online Contents | 2010
|COMPOSITES: A 'super scanner' for nondestructive testing
Online Contents | 1994
|