摘要 |
PROBLEM TO BE SOLVED: To provide a data generation device in which character information is extracted out of a printed matter printed in plural colors.SOLUTION: A data generation device separates image data generated by reading an advertising catalog in color by respective colors to generate color image data for each of a plurality of colors; and stores, when an analysis area including a character area, which is recognized as a character, is detected out of each of the generated color image data, word information made of character information detected by analyzing the character area included in the analysis area, and positional information indicative of coordinates in the image data in the character area in association with each other in an analysis character information storage part. When a retrieval keyword is input, the data generation device generates a similar retrieval keyword obtained by replacing a character included in the retrieval keyword with another character similar to the character, retrieves the positional information out of the analysis character information storage part on the basis of the retrieval keyword and the similar retrieval keyword, and outputs the retrieved positional information. |