{"created":"2025-01-19T01:19:38.553553+00:00","updated":"2025-01-19T14:50:51.515723+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00219570","sets":["1164:5251:10854:11006"]},"path":["11006"],"owner":"44499","recid":"219570","title":["船上監視を支援する物体検出システムの設計"],"pubdate":{"attribute_name":"公開日","attribute_value":"2022-08-25"},"_buckets":{"deposit":"64a3b3c0-97b5-4141-a2ce-b608ed9008b7"},"_deposit":{"id":"219570","pid":{"type":"depid","value":"219570","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"船上監視を支援する物体検出システムの設計","author_link":["572793","572794","572792"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"船上監視を支援する物体検出システムの設計"},{"subitem_title":"Design of Object Detection System to Support Maritime Surveillance","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"センシング・検出（DPS）","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2022-08-25","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"総合研究大学院大学情報学専攻"},{"subitem_text_value":"日立製作所"},{"subitem_text_value":"国立情報学研究所／総合研究大学院大学情報学専攻"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Department of Informatics, The Graduate University for Advanced Studies, SOKENDAI","subitem_text_language":"en"},{"subitem_text_value":"Hitachi, Ltd.","subitem_text_language":"en"},{"subitem_text_value":"National Institute of Informatics / Department of Informatics, The Graduate University for Advanced Studies, SOKENDAI","subitem_text_language":"en"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_publisher":{"attribute_name":"出版者","attribute_value_mlt":[{"subitem_publisher":"情報処理学会","subitem_publisher_language":"ja"}]},"publish_status":"0","weko_shared_id":-1,"item_file_price":{"attribute_name":"Billing file","attribute_type":"file","attribute_value_mlt":[{"url":{"url":"https://ipsj.ixsq.nii.ac.jp/record/219570/files/IPSJ-EIP22097013.pdf","label":"IPSJ-EIP22097013.pdf"},"date":[{"dateType":"Available","dateValue":"2024-08-25"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-EIP22097013.pdf","filesize":[{"value":"1.4 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"660","billingrole":"5"},{"tax":["include_tax"],"price":"330","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"26"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"6884aefc-f902-4359-9912-48834bdaaff1","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2022 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"陳, 明康"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"孫, 静涛"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"合田, 憲人"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA11238429","subitem_source_identifier_type":"NCID"}]},"item_4_textarea_12":{"attribute_name":"Notice","attribute_value_mlt":[{"subitem_textarea_value":"SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc."}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_18gh","resourcetype":"technical report"}]},"item_4_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"2188-8647","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"船上監視をはじめとする船舶操縦者支援のため，画像認識技術を用いた海上の物体検出技術の開発が期待されている．AI を用いた画像認識技術の進展により，海上の船舶等の物体を検出することが可能となってきたが，実際の航行では，既存技術による物体検出精度が著しく低下するという問題がある．この問題を引き起こす原因は二つがあり，一つ目は雨などの天候のノイズによる画像品質劣化であり，二つ目は大規模な学習データセットの欠如である．本稿では，船上監視を支援するための物体検出システムの設計について述べる．本システムでは，船上で撮影される動画像に対して天候によるノイズを除去した上で，AI を用いた物体検出を実行することにより，画像品質を改善し，精度の高い物体検出を可能とする．また，航行中に撮影される様々なシーンの動画像データを船外のクラウドに収集し，物体検出モデルを再学習することにより，モデルの更新を行う．最後に，天候ノイズによる検出精度への影響とノイズ除去の画質向上効果を検証し，提案システムのプロトタイプ実装を用いた予備評価結果についても報告する．","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"The development of maritime object detection technology based on image recognition technology is designed to help ship operators in maritime surveillance and other areas. However, in actual navigation, the accuracy of object detection is degraded significantly with conventional approaches. There are two primary causes of this problem: first, image quality degradation owing to noise induced by weather conditions such as rain, and second, a lack of large training datasets for maritime surveillance. This paper describes the design of an object detection system for maritime surveillance. By performing AI-based object detection on video images recorded on board after removing noise caused by weather conditions, the system improves image quality and allows highly accurate object detection. The AI model for object detection is updated by collecting video data of various scenes in a cloud outside the ship and re-training the object detection model. Finally, we report on the impact of weather noise on detection accuracy and the effect of noise removal on image quality, as well as preliminary evaluation results using a prototype implementation of the proposed system.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"6","bibliographic_titles":[{"bibliographic_title":"研究報告電子化知的財産・社会基盤（EIP）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2022-08-25","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"13","bibliographicVolumeNumber":"2022-EIP-97"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"id":219570,"links":{}}