| Presentation | 2022-01-28 Proposal of mesh abnormality detection method in knitting machine using AI Shoya Ogawa, Nobuyuki Yonaga, Kazuki Fukae, Tetsuro Imai, Kenichi Arai, Toru Kobayashi, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Highly durable tortoiseshell nets are attracting attention as fishing nets and rockfall prevention nets for slopes. These nets are manufactured by a machine called a knitting machine, and high quality nets are manufactured by constantly monitoring the mesh for abnormalities. However, it is difficult to distinguish the abnormal meshes from the normal meshes, and it relies on the intuition and experience of skilled technicians. In this study, we propose a method for detecting abnormal meshes in knitting machines using image analysis AI. In this method, we focus on the twisted part of the yarn that constitutes the mesh, extract the shape of the twisted part as a feature, and classify it by SVM. In the case of image classification by deep learning, it is necessary to prepare a large amount of training data because the target image contains a lot of noise that varies from day to day, such as water vapor in the factory or dirt on the rollers. By applying our method to an actual knitting machine, we confirmed that our method requires less training data than image classification by deep learning, and that it can detect abnormalities in real time. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | AI / Machine Learning / SVM / CNN / Image Recognition / IoT |
| Paper # | ICM2021-40,LOIS2021-38 |
| Date of Issue | 2022-01-20 (ICM, LOIS) |
| Conference Information | |
| Committee | LOIS / ICM |
|---|---|
| Conference Date | 2022/1/27(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Online |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Practical Use of Lifelog, Office Information System, Business Management, etc. |
| Chair | Toru Kobayashi(Nagasaki Univ.) / Kazuhiko Kinoshita(Tokushima Univ.) |
| Vice Chair | Hiroyuki Toda(NTT) / Haruo Ooishi(NTT) / Eiji Takahashi(NEC) |
| Secretary | Hiroyuki Toda(Nagasaki Univ.) / Haruo Ooishi(NTT) / Eiji Takahashi(Bosco) |
| Assistant | Kazuki Fukae(Nagasaki Univ.) / Yoshifumi Kato(NTT) |
| Paper Information | |
| Registration To | Technical Committee on Life Intelligence and Office Information Systems / Technical Committee on Information and Communication Management |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Proposal of mesh abnormality detection method in knitting machine using AI |
| Sub Title (in English) | |
| Keyword(1) | AI |
| Keyword(2) | Machine Learning |
| Keyword(3) | SVM |
| Keyword(4) | CNN |
| Keyword(5) | Image Recognition |
| Keyword(6) | IoT |
| 1st Author's Name | Shoya Ogawa |
| 1st Author's Affiliation | Nagasaki University(Nagasaki Univ.) |
| 2nd Author's Name | Nobuyuki Yonaga |
| 2nd Author's Affiliation | Nagasaki University(Nagasaki Univ.) |
| 3rd Author's Name | Kazuki Fukae |
| 3rd Author's Affiliation | Nagasaki University(Nagasaki Univ.) |
| 4th Author's Name | Tetsuro Imai |
| 4th Author's Affiliation | Hiroshima City University(Hiroshima City Univ.) |
| 5th Author's Name | Kenichi Arai |
| 5th Author's Affiliation | Nagasaki University(Nagasaki Univ.) |
| 6th Author's Name | Toru Kobayashi |
| 6th Author's Affiliation | Nagasaki University(Nagasaki Univ.) |
| Date | 2022-01-28 |
| Paper # | ICM2021-40,LOIS2021-38 |
| Volume (vol) | vol.121 |
| Number (no) | ICM-354,LOIS-355 |
| Page | pp.pp.40-45(ICM), pp.40-45(LOIS), |
| #Pages | 6 |
| Date of Issue | 2022-01-20 (ICM, LOIS) |