摘要 |
An image processing apparatus separates in a scanned image a text area from a graphic area primarily including a graphic form or a graph. For the text area, neighboring black pixels are connected to perform character determination in a unit of a rectangle obtained by connecting the black pixels. For the graphic area, labeling processing is used to extract a circumscribed rectangle of consecutive black pixels, without connecting the black pixels, to perform character determination in a unit of the circumscribed rectangle.
|