| 摘要 |
PROBLEM TO BE SOLVED: To make it possible to identify blocks whose contents are continued to each other by judging the connection of character string areas extracted from an input image as respectively different blocks by using morpheme analysis. SOLUTION: Data of continuous blocks are acquired (steps 30-4, 30-6), the connection of character strings in these blocks is judged by morpheme analysis (step 30-7), and at the time of judging that these character strings are connected to each other, those character strings are continuously stored in the same character storing area (step 30-9). Since character strings are stored by inserting a paragraph shift code as necessary, the setting of paragraphs is utilized also for succeeding edition processing and an output format can be maintained. |
