compareWord: Compares OCR words to truth

In duncantl/Rtesseract: Interface to the tesseract OCR system

Compares OCR words to truth

Description

If we have the true text, we can use compareWords and other functions to compare the OCR results to the truth and determine which symbols were matched incorrectly. The results can then be displayed on the image and the incorrect symblols identified.

compareWords process a collection of words; compareWordInfo processes a single word and the corresponding true/actual value for the word.

Usage

compareWords(ocr, truth)
compareWordInfo(ocr, truth)

Arguments

ocr	the words from the OCR classification
truth	the true words

Value

compareWords returns a data frame with a row for each symbol/character that was different between the OCR version and the truth. The data frame contains

ocr	the character recognized by the OCR system, incorrectly
truth	the true value of the character
position	the index in the word of the misclassified character/symbol
wordIndex	the index of the word in which the misclassification occured
trueWord	the value of the true word
ocrWord	the value of the word as recognized by the OCR system
symbolIndex	the index of the character/symbol in the entire set of symbols

Note

This function does not yet handle the case where the OCR and true words do not have the same length.

Author(s)

Duncan Temple Lang

References

Tesseract https://code.google.com/p/tesseract-ocr/

See Also

tesseract, GetText, GetConfidences

Examples

compareWords(c("Duncin", "Temple", "Lung"), c("Duncan", "Temple", "Lang"))

duncantl/Rtesseract documentation built on Sept. 8, 2024, 8:38 a.m.