| Presentation | 2020-03-16 Metal Material Analysis using SEM Images of Multiple Magnifications Ryunosuke Ito, Keisuke Kameyama, Hideitsu Hino, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In recent years, in order to quickly develop metal materials having special properties, it is urgently necessary to make development and verification more efficient. In particular, prediction of properties using material images has a wide range of applications and is attracting attention. Surface images of metal materials can be obtained by a scanning electron microscope at a magnification of about several hundred to several hundred thousand.However, using only material images of a single magnification may cause a drawback in analysis. Therefore, in order to solve the problem by combining effective features for each scale, many studies have been conducted on a prediction method that can use a combination of a plurality of magnifications. In this work, we propose the selective forked autoencoder (SFAE). The SFAE is a neural network architecture in which multiple autoencoders are connected in parallel, and trained to perform so that a specific autoencoder responds to a specific pattern input. We applied the proposed architecture to a SEM image of the actual metal material surface, and compared the detection accuracy of the striation, which indicates metal fatigue, with the detection method using histogram. It is confirmed that the proposed architecture has higher detection performance than the classical patch classification method and the classifier based on a simple neural network. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | material analysis / SEM / striation / multiscale / autoencoder |
| Paper # | PRMU2019-78 |
| Date of Issue | 2020-03-09 (PRMU) |
| Conference Information | |
| Committee | PRMU / IPSJ-CVIM |
|---|---|
| Conference Date | 2020/3/16(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | Yoichi Sato(Univ. of Tokyo) |
| Vice Chair | Toru Tamaki(Hiroshima Univ.) / Akisato Kimura(NTT) |
| Secretary | Toru Tamaki(NTT) / Akisato Kimura(OMRON SINICX) |
| Assistant | Yusuke Uchida(DeNA) / Takayoshi Yamashita(Chubu Univ.) |
| Paper Information | |
| Registration To | Technical Committee on Pattern Recognition and Media Understanding / Special Interest Group on Computer Vision and Image Media |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Metal Material Analysis using SEM Images of Multiple Magnifications |
| Sub Title (in English) | |
| Keyword(1) | material analysis |
| Keyword(2) | SEM |
| Keyword(3) | striation |
| Keyword(4) | multiscale |
| Keyword(5) | autoencoder |
| 1st Author's Name | Ryunosuke Ito |
| 1st Author's Affiliation | University of Tsukuba(Univ. Tsukuba) |
| 2nd Author's Name | Keisuke Kameyama |
| 2nd Author's Affiliation | University of Tsukuba(Univ. Tsukuba) |
| 3rd Author's Name | Hideitsu Hino |
| 3rd Author's Affiliation | The Institute of Statistical Mathematics(ISM) |
| Date | 2020-03-16 |
| Paper # | PRMU2019-78 |
| Volume (vol) | vol.119 |
| Number (no) | PRMU-481 |
| Page | pp.pp.71-76(PRMU), |
| #Pages | 6 |
| Date of Issue | 2020-03-09 (PRMU) |