WEKO3
アイテム
Importance of input image size on the performance of automatic age determination of chum salmon Oncorhynchus keta using deep learning
https://fra.repo.nii.ac.jp/records/2015206
https://fra.repo.nii.ac.jp/records/20152060bf309df-c332-4731-a3ae-0a97c36b40aa
| Item type | 学術雑誌論文 / Journal Article(1) | |||||
|---|---|---|---|---|---|---|
| 公開日 | 2025-09-29 | |||||
| タイトル | ||||||
| タイトル | Importance of input image size on the performance of automatic age determination of chum salmon Oncorhynchus keta using deep learning | |||||
| 言語 | en | |||||
| 言語 | ||||||
| 言語 | eng | |||||
| キーワード | ||||||
| 言語 | en | |||||
| 主題Scheme | Other | |||||
| 主題 | Age reading; Artificial intelligence; Deep learning; Fine-tuning; Fish scales; Grad-CAM; ResNet; Salmonid | |||||
| 資源タイプ | ||||||
| 資源タイプ識別子 | http://purl.org/coar/resource_type/c_6501 | |||||
| 資源タイプ | journal article | |||||
| 著者 |
多賀, 悠子
× 多賀, 悠子× 鈴木, 健吾× 高橋, 昌也× 平林, 幸弘× 大井, 邦昭× 井上, 誠章 |
|||||
| 抄録 | ||||||
| 内容記述タイプ | Abstract | |||||
| 内容記述 | Manual determination of fish age using hard tissues demands considerable effort. However, reliable and high-performance automatic alternative methods are not widely available. The aim of this study was to investigate the influence of input image size on the performance and the image areas used for automatic age determination. To this end, we used input images with sizes of 240 × 240–960 × 960 or 170 × 340–679 × 1358 pixels, either whole or trimmed in half, of 3- to 5-year-old chum salmon scales (n = 1835), as well as deep convolutional neural networks (CNN). In entire-scale images, high accuracy was achieved when the input image size exceeded 679 × 679 pixels, reaching a maximum of 94.7%. Below this size, accuracy decreased significantly, and overfitting became pronounced. At sizes exceeding 480 × 480 pixels, the CNN consistently based its determinations on areas outside the first annulus, similar to visual inspection. Conversely, at sizes below 480 × 480 pixels, where the circuli become indistinguishable, the CNN focused on a wide range around the focus. In trimmed-scale images, CNN accuracy plateaued at a lower level (80.8–88.7%) than in entire-scale images, even for large-sized images. These results suggest that using sufficiently large entire-scale images which retain detailed information about the circulus pattern is important for achieving high performance of frameworks for automatic age determination. | |||||
| 言語 | en | |||||
| 書誌情報 |
en : Fisheries Science 巻 91, 号 6, 発行日 2025-09-16 |
|||||
| 出版者 | ||||||
| 出版者 | Springer | |||||
| 言語 | en | |||||
| 出版者 | ||||||
| 出版者 | 日本水産学会 | |||||
| 言語 | ja | |||||
| ISSN | ||||||
| 収録物識別子タイプ | PISSN | |||||
| 収録物識別子 | 0919-9268 | |||||
| 書誌レコードID | ||||||
| 収録物識別子タイプ | EISSN | |||||
| 収録物識別子 | 1444-2906 | |||||
| DOI | ||||||
| 関連タイプ | isIdenticalTo | |||||
| 識別子タイプ | DOI | |||||
| 関連識別子 | 10.1007/s12562-025-01917-y | |||||
| 情報源 | ||||||
| 関連タイプ | isIdenticalTo | |||||
| 識別子タイプ | Local | |||||
| 関連識別子 | 25341001 | |||||
| 言語 | ja | |||||
| 関連名称 | 水産技術研究所 環境・応用部門 水産工学部 （神栖） | |||||
| 関連サイト | ||||||
| 関連タイプ | isIdenticalTo | |||||
| 識別子タイプ | URI | |||||
| 関連識別子 | https://link.springer.com/article/10.1007/s12562-025-01917-y | |||||
| 言語 | en | |||||
| 関連名称 | SPRINGER NATURE Link | |||||
| 著者版フラグ | ||||||
| 出版タイプ | SMUR | |||||
| 出版タイプResource | http://purl.org/coar/version/c_71e4c1898caa6e32 | |||||
