Back to FabImage Studio website

You are here: Start » Filter Reference » Computer Vision » Deep Learning » MergeCharactersIntoLines

MergeCharactersIntoLines

Expert

Module: DL_OCR

Converts a output of DL_ReadCharacters to lines of text.

Name Type Range Description
Input value inCharacters OcrResultArray Output of DL_ReadCharacters
Input value inMaxGap Real 0.0 - 10.0 Maximum horizontal gap between joint characters' boxes, denoted as fraction of 'A' char height
Input value inMaxShift Real 0.0 - 1.0 Maximum vertical misalignment between joint character's boxes, denoted as fraction of 'A' char height
Input value inMargin Real 0.0 - 10.0 Additional margin added to result, denoted as fraction of 'A' char height
Input value inMinLength Integer 1 - 200 Minimal number of chars to create line
Input value inFlatten Bool If True, it concatenates the words on the line into a single result string, otherwise each word is a separate result string
Output value outLines Rectangle2DArray Minimal Box which cover all selected character boxes
Output value outStrings StringArray Text of merged characters
Output value outMapping Integer?Array Mapping between input characters and output lines, outMapping[i] stores the index line to which inCharacters[i] belongs. If outMapping[i] is NIL it means that inCharacters[i] has not been added to any line

Description

This tool takes the text contained in OcrResultArray from the DL_ReadCharacters and merges it into lines.

Hints

  1. Depending on the inMaxGap and inMaxShift values, we can get different number of lines. See the image below, where increasing the inMaxGap results in one line of text, whereas a smaller value will return two separate lines:

    2. Different values of the inMaxGap and inMaxShift parameters result in a different number of lines.
  2. The lines are sorted by the Y value, e.g.:


  3. The tool can also be used to get rid of false characters by setting a different value of the inMinLength parameter. In the image below setting the inMinLength to 2 resulted in filtering out single false characters returned by the DL_ReadCharacters tool.


Complexity Level

This filter is available on Expert Complexity Level.

See Also

  • DL_ReadCharacters – Performs optical character recognition using a pretrained deep learning model.