PDF
Full
Outline
Digital Image Correlation (DIC) is a non-contact optical measurement technique that uses speckle patterns as deformation carriers to measure surface displacement and deformation fields of objects. It has been widely applied in key industrial fields such as aerospace, mechanical engineering, and power engineering. In general, specialized software is required for Digital Image Correlation (DIC) measurement and analysis. In particular, in the measurement of fatigue and dynamic problems, it is essential to address challenges arising from big data processing, such as long computation times and low efficiency. With the development of artificial intelligence technology, deep learning provides new opportunities for DIC method. However, a huge dataset is required for the construction of DIC deep learning network, which not only increases the cost of data collection but also takes a long computation time. To solve the above problems, this paper proposes a DIC-2D displacement measurement method based on migration learning, which is based on U-Net network including a multi-level feature extractor, an attention mechanism and a depth-separable convolution. In the pre-training process of the network, simulated scattering images are used as the training dataset to form the pre-trained network；On this basis, multiple transfer learning fine-tuning strategies are used to optimize the network parameters using a small number of real speckle images with different mean intensity gradients to establish the migration network, and real speckle images are used for verification. The analysis results show that the network trained by the global fine-tuning strategy exhibits higher accuracy and better robustness in the training of different mean intensity gradient speckle images. The DIC migration learning method proposed in this paper can significantly reduce the training time and cost for data acquisition.
PDF
128
61
| 科 Family | 属数 Number of genus | 种数 Number of species | 占总种数比例 Percentage of total species (%) | 属 Genus | 种数 Number of species | 占总种数比例 Percentage of total species (%) |
|---|---|---|---|---|---|---|
| 鹅膏菌科Amanitaceae | 2 | 11 | 5.26 | 鹅膏菌属 Amanita | 10 | 4.78 |
| 小菇科 Mycenaceae | 2 | 12 | 5.74 | 丝盖伞属 Inocybe | 5 | 2.39 |
| 多孔菌科 Polyporaceae | 8 | 14 | 6.70 | 蜡蘑属 Laccaria | 5 | 2.39 |
| 红菇科 Russulaceae | 3 | 23 | 11.00 | 小皮伞属 Marasmius | 6 | 2.87 |
| 小菇属 Mycena | 11 | 5.26 | ||||
| 光柄菇属 Pluteus | 5 | 2.39 | ||||
| 红菇属 Russula | 17 | 8.13 | ||||
| 栓菌属 Trametes | 5 | 2.39 |
关闭全屏
No. 86 Xueyuan South Road, Haidian District, Beijing
010-62199257
qkjq@cast.org.cn
Copyright © 2025 China Association for Science and Technology. All rights reserved. For all open access content, the relevant licensing terms apply.
Sponsored by the Office of the Leading Group for Cybersecurity and Informatization of CAST, and supported by Science and Technology Review Publishing House
