A Novel Thinning Scheme for Reducing Thinning Distortions

概要

論文の詳細を見る
Thinning is one of the fundamental binary image processing methods. It deletes some dark points marked with value'1'in binary image and transforms the pattern into thin representation without changing its connectivity properties. This processing is used frequently in character recognition and understanding line drawings. This paper describes a novel thinning scheme for reducing thinning distortions. The proposed thinning algorithms consist of two main procedures : thinning and recovering. The thinning procedure searches for the deletable candidates satisfying thinning conditions. The recovering procedure is used to check the distribution of deletable candidates resulted from the thinning procedure in the same iteration and to recover some of them which may cause distortions into undeletable ones.
一般社団法人情報処理学会の論文
1990-03-14