Image retrieval based on character-matching technology

摘要：

In order to introduce effective text-retrieval technologies into content-based image retrieval, a novel method based on string-matching technology was presented incorporating the different sensitivity along variant directions of human visual system. The image is firstly divided into non-overlapped blocks and sub-feature is extracted according to the human visual feature. Then, a novel descriptor is defined. Based on which, the sub-features is mapping into characters and the image into the text. Finally, the technology of the character string matching is adopted as the similarity measurement. Meantime, the situation information of character is considered, an adjusted method is proposed. Experimental results show that the proposed method has sound and robust retrieval performance.

作者: Zhao Shan Liu Jing

作者单位: School of Computer Science and technology Henan Polytechnic University Jiaozuo 454003 China;Jiangsu School of Computer Science and technology Henan Polytechnic University Jiaozuo 454003 China

会议类型: 国际会议

会议名称: The Fifth International Conference on Information Assurance and Security(第五届信息保障与安全国际会议)

会议地点: 西安

会议语种:英文

页码: 369-372

在线出版日期: 2009-08-18（万方平台首次上网日期，不代表论文的发表时间）