| 摘要 |
A method for converting an image includes the steps of: obtaining an inclination of a character in the image; obtaining apex pixels forming contours of the character; obtaining an identical line segment candidate pixel group from a portion supposed to be expressed by a single line segment; extracting, based on the inclination, the identical line segment candidate pixel group as an identical line segment pixel group; obtaining both-end pixels which are apex pixels located on both sides of the identical line segment pixel group thus extracted; approximating, by a line segment connecting between the both-end pixels thus obtained, a contour of the character in a portion corresponding to the both-end pixels; and approximating a portion, excluding the portion that has been approximated, by a straight line or a curved line by performing an outlining process using apex pixels other than the apex pixels of the identical line segment pixel group.
|
