論理和型および信頼区間型の探索領域拡張による複数関心パタンの順序探索
論理和型および信頼区間型の探索領域拡張による複数関心パタンの順序探索
カテゴリ: 論文誌(論文単位)
グループ名: 【C】電子・情報・システム部門
発行日: 2022/08/01
タイトル(英語): Ordered Search for Multiple Feature Patterns Based on Area Expansion by Union Operation and Confidential Interval
著者名: 浅野 裕一(鎌倉光機(株)/北海道大学),高氏 秀則(北海学園大学),田中 孝之(北海道大学),金子 俊一(北海道大学)
著者名(英語): Hirokazu Asano (Kamakura Koki Co., Ltd./Hokkaido University), Hidenori Takauji (Hokkai-Gakuen University), Takayuki Tanaka (Hokkaido University), Shun'ichi Kaneko (Hokkaido University)
キーワード: 画像処理,順序探索,論理和,信頼区間,領域拡張_x000D_ image processing,order search,union,confidential interval,region expansion
要約(英語): Two novel strategies for incremental searching of interest patterns are proposed in context of sequential search of multiple patterns in the scene. One is based on a scheme based on union operation to extend search regions depending on all of the proceeding searches, which can possibly adapt the search to some local distortion in the scene or the objects. Another strategy utilizes incremental setting of confidence intervals to the search regions depending on their statistics in position errors of respective patterns. In addition to them, a new scheme of repetitive trials for realizing reliability in the search. By using a set of real objects, many experiments show the effectivity and the efficiency of the proposed strategies.
本誌掲載ページ: 971-981 p
原稿種別: 論文/日本語
電子版へのリンク: https://www.jstage.jst.go.jp/article/ieejeiss/142/8/142_971/_article/-char/ja/
受取状況を読み込めませんでした
