| Autor: |
Guodong Liu, Fupeng Wei, Fengdong Chen, Zhitao Peng, Jun Tang, Bingguo Liu, Qihua Zhu, Dongxia Hu, Bin Feng, Yong Xiang, Chengcheng Wang |
| Jazyk: |
angličtina |
| Rok vydání: |
2018 |
| Předmět: |
|
| Zdroj: |
IEEE Access, Vol 6, Pp 30402-30413 (2018) |
| Druh dokumentu: |
article |
| ISSN: |
2169-3536 |
| DOI: |
10.1109/ACCESS.2018.2843759 |
| Popis: |
In this paper, a hierarchical kernel extreme learning machine (HK-ELM) is proposed to solve the problem that the size of laser-induced damage (LID) cannot be accurately measured under the condition of inhomogeneous total internal reflection illumination. The proposed method overcomes the limitations of the radiometric method, which depends on a single factor to measure the size of damage. After analyzing the light scattering characteristics of LID with the finite-difference time-domain method, we find that we should consider the influences of multiple factors when measuring the size of damage under the condition of uneven illumination. The experimental results show that HK-ELM can achieve ultra-resolution measurements, and the mean relative error of the measured size predicted by HK-ELM is within 5% on the testing samples. Compared with the traditional technology, the method proposed in this paper effectively improves the measurement accuracy. |
| Databáze: |
Directory of Open Access Journals |
| Externí odkaz: |
|
