A Block Growing Approach to the Generation of Reading Order of Chinese Newspapers

• Main Authors: King, Chi-Faang, 金繼昉
• Other Authors: Fan, Kuo-Chin
• Format: Others
• Language: zh-TW
• Published: 1996
• Online Access: http://ndltd.ncl.edu.tw/handle/02629075175385504768

碩士 === 國立中央大學 === 資訊工程研究所 === 84 === 　　The reading order of articles plays an important role in the field of document understanding. Since the reading sequence conveys the semantic information of documents, it will influence the correctness of the post-applications in document analysis. Hence, the reading order generation of documents which most researtchers and systems frequently ignore is an indispensable post-processing in document understanding systems. 　　In this thesis, a block growing approach is proposed to generate the reading order of Chinese newspapers by matching the preddfined style graphs. In this graph matching approach, the geometric rdlation graph (GRG) is first constructed according to the geometric relations among the segmented blocks of input documents. Those blocks belonging to an article are gradually merged together to grow up an article by matching the predefined style graphs. Since the local reading information is kept during the block growing processes, the global reading information can be easily generated. This bottom-up merging method provides a powerful and flexible approach in finding the reading order of articles. 　　Some Chinese newspapers with horizontal and vertical styles are tested to verify the feasibility of our proposed method. Experimental results reveal the effectiveness of our proposed approach.