| 摘要 |
PROBLEM TO BE SOLVED: To improve the accuracy of recognizing a character in a document.SOLUTION: An information processing apparatus 10 for recognizing a character in a character recognition area in a document specifies a first vertex 21a from a document image 21 acquired from the document, extracts a first distortion amount indicating a degree of distortion of the first vertex 21a, detects a reference mark 21c arranged at a corner 21b including the first vertex 21a, to be used for correcting printing misregistration of the character recognition area, from the document image 21, extracts a second distortion amount indicating a degree of distortion of the reference mark 21c, generates a corrected document image by correcting the distortion of the first vertex 21a from the document image 21, and acquires a first correction amount of the first vertex 21a in correcting the document image 21 into the corrected document image. The information processing apparatus 10 calculates a second correction amount of the reference mark 21c, on the basis of the first distortion amount, the first correction amount, and the second distortion amount, to identify a position of the reference mark 21c in the corrected document image, on the basis of the second correction amount.SELECTED DRAWING: Figure 1 |
