A retrieval method of ancient Chinese character images based on neighbor clustering

Abstract

Character image clustering is an assistant way of studying ancient Chinese characters. This paper analyzed the feature of ancient Chinese characters. Aiming at the key problems such as feature extraction and clustering, we designed a simple and effective neighbor clustering method for ancient Chinese character images which extracts adoptive feature of the coarse periphery features, the second coarse periphery features and the edge directional line element feature of ancient characters images. The experimental results show that the character clustering algorithm is effective.