The OCR (Optical character recognition) module is used for finding individual characters or words in the image.
You can define ROIs (regions of interest) - areas in which the characters/words are supposed to be found. ROIs can then also be used in the evaluation to define specific rules for them.
The detection threshold is a threshold for the text detector. As it goes higher, less text is being detected and vice versa.
Word completeness threshold
The word completeness threshold is a threshold for detecting letters or words. If the threshold is small, you get bounding boxes around whole words, whereas if you put it higher, you get bounding boxes around smaller parts of words or individual letters.
The omit threshold is a threshold for text recognition. If the character/word is recognized with lower confidence than this threshold, it will not be displayed (works the same way as confidence for Detector).
Evaluation is done based on regular expressions (regex) - each found word/letter is tested, whether it matches the regex. If you defined ROIs, you can set specific rules for each of them or general rules for characters/words in any of them.