Read/Search this Article
Abstract
選択的注意のモデルは,複数のカテゴリーのパターンが組み合わさった複合パターンに対し個々のパターンを認識し,セグメンテーションを行なうシステムである.従来,このシステムを漢字認識に応用したモデルの研究がされてきたが,その際にセグメンテーションの不完全さが指摘されていた.本論文では,直線の内挿過程と呼ぶ過程を導入したモデルによって,この問題点の克服を試みた.
The "selective attention model" proposed by Fukushima has a large ability to recognize composite figures consiting of many patterns. Previously, several works have been done to use this model to recognize Chinese Characters, but the ability of their systems to recognize distored patterns was not large enough. We propose a "line interpolation process" and apply it to the selective attension model to improve the ability of segmentation.
Journal
- IEICE technical report. Neurocomputing [List of Volumes]
-
IEICE technical report. Neurocomputing 95(189), 151-158, 1995-07-27 [Table of Contents]
The Institute of Electronics, Information and Communication Engineers