{"created":"2025-01-18T23:17:19.906554+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00053171","sets":["1164:4619:4695:4697"]},"path":["4697"],"owner":"1","recid":"53171","title":["複数のGMRFモデルを用いた屋外シーンの領域分割"],"pubdate":{"attribute_name":"公開日","attribute_value":"1996-09-12"},"_buckets":{"deposit":"1e4f6137-774a-41d1-827f-776f0e3d6fff"},"_deposit":{"id":"53171","pid":{"type":"depid","value":"53171","revision_id":0},"owners":[1],"status":"published","created_by":1},"item_title":"複数のGMRFモデルを用いた屋外シーンの領域分割","author_link":["0","0"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"複数のGMRFモデルを用いた屋外シーンの領域分割"},{"subitem_title":"Unsupervised Segmentaion of Outdoor Scenes by using Multiple GMRF Models","subitem_title_language":"en"}]},"item_type_id":"4","publish_date":"1996-09-12","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"日本アイ・ビー・エム（株）東京基礎研究所"},{"subitem_text_value":"郵政省通信総合研究所"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"IBM Research, Tokyo Research Laboratry","subitem_text_language":"en"},{"subitem_text_value":"Commumication Research Laboratory Ministry of Posts and Telecommunications","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/53171/files/IPSJ-CVIM96101001.pdf"},"date":[{"dateType":"Available","dateValue":"1998-09-12"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-CVIM96101001.pdf","filesize":[{"value":"663.3 kB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"660","billingrole":"5"},{"tax":["include_tax"],"price":"330","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"20"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"1f7c84c8-6c9b-4377-8f91-55aaf41b4dbe","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 1996 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"越後, 富夫"},{"creatorName":"飯作俊一"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Tomio, Echigo","creatorNameLang":"en"},{"creatorName":"Shun-Ichi, Iisaku","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA11131797","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_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"屋外の自然画像にはテクスチャ領域が多く含まれる．符号化や画像理解の目的には，テクスチャ領域を色だけでなく，テクスチャ特徴を利用して，特徴毎に領域を切り出すことが重要である．テクスチャ特徴を表現するには，画素単位ではなく，領域画像を必要とするため，領域境界近傍でテクスチャ特徴を利用することが困難であった．本手法では，ガウシアンマルコフ確率場(MR)でモデル化し，複数の定義パラメータを使い分け，領域境界で併合初期の小領域にもテクスチャ特徴を表現するGMRFパラメータを使った領域併合を行なう．さらに併合領域が拡大すると，より信頼度の高いGMRFパラメータで領域併合を行なう．以上のようにして，なめらかな境界線で画像を分割し，かつ，境界線近傍までテクスチャ特徴で領域併合を行なう手法を提案する．","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"The goal of this research is to make region-based coding method of images effective. Texture regions implied in outdoor scene should be segmented by not only colors, but also textured features. Gaussian Markov Random Fields (GMRF) represent texture efficiently, however, the parameters of GMRF requires a large regions for exactness, which means GMRF is not available in neighborhood of region boundary. Multiple models of GMRF can merge small regions in neighborhood of region boundary and at the initial merging steps, because a few parameters require small regions and regions merged reliably by more parameters can be obtained from gathering small regions.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"8","bibliographic_titles":[{"bibliographic_title":"情報処理学会研究報告コンピュータビジョンとイメージメディア（CVIM）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"1996-09-12","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"86(1996-CVIM-101)","bibliographicVolumeNumber":"1996"}]},"relation_version_is_last":true,"weko_creator_id":"1"},"links":{},"id":53171,"updated":"2025-01-22T06:26:13.471399+00:00"}