By using the API of the character recognition library, you can provide users with various methods of specifying text. By specifying straight lines or rectangles, you can easily specify text with any inclination degree. In addition, there are many dedicated APIs available, such as extracting paragraphs within polygons and specifying text excluding surrounding line shapes.

Text specification methods

1. Extract multiple paragraphs collectively by setting one parameter.
   
2. Extract multiple paragraphs in multiple stages by setting multiple parameters.
   For example, extract in three stages using parameters such as horizontal writing, vertical writing, and symbols.
   
3. Extract multiple paragraphs collectively within rectangles, inclined rectangles, or polygons.
   
4. Extract one-line text by specifying line segments.
   
5. Extract one-paragraph text using rectangles or inclined rectangles as hooks.
   Text that falls within rectangles and the frame of rectangles.
   
6. Extract one-paragraph text using the inner part of rectangles or inclined rectangles.
   Text within rectangles (excluding those that fall on the frame).
   
7. Force specification of text within rectangles or inclined rectangles.
   It is possible to specify even text that intersects with parts of shapes or lines.
   
8. Specify one character within rectangles or inclined rectangles.
   Specify the interior of rectangles as one character.