Department News

[MEch-SSENGER] Development of a method for estimating non-acoustic properties using micro-CT images

Author
관리자
Date
2021-09-13
Views
462


 

Professor Yeon June Kang's research team developed a method for estimating non-acoustic properties such as the porosity and curvature of fibrous acoustic materials using Micro-CT images.

The non-acoustic properties of fiber-type acoustic materials are considered as important factors in estimating the acoustic performance of the material. However, measuring it is a task that requires a lot of time and trained workers who can handle the measuring equipment proficiently. To solve this problem, the research team developed a convolutional neural network that can estimate non-acoustic properties through Micro-CT images and obtained about 1500 Micro-CT images that can be used for subsequent research. This research is expected to reduce the cost of finding non-acoustic properties and to lay the groundwork for the development of a new research field that combines neural networks with traditional acoustic materials.

Jeon, Ju Hyun, et al. "Convolutional neural networks for estimating transport parameters of fibrous materials based micro-computerized tomography images." The Journal of the Acoustical Society of America 149.4 (2021): 2813-2828.

MEch-SSENGER  JUNG Guyeop