My Learnings – OCR in SAP IRPA

As explained by Abinaya Seenivasan in her blog https://blogs.sap.com/2020/02/23/how-to-build-custom-ocr-in-sap-rpa/ about OCR, I have followed similar steps and wanted to share few steps which I have felt missing there along with understanding and learnings.

1. Download Tesseract OCR from below link for Windows

https://github.com/UB-Mannheim/tesseract/wiki

2. Update Training data from Tesseract Best data

https://github.com/tesseract-ocr/tessdata_best

3. Simple Python Script required to execute OCR

import pytesseract
import json
import os

# provide the file path where all your files located
filePath = 'D:\\IRPA_TEMP\\'

#Set location where Tessearct is installed
pytesseract.pytesseract.tesseract_cmd = r'C:\Program Files\Tesseract-OCR\tesseract.exe'

#Consider all files from the folder
files = os.listdir(filePath)

#Considering file is already in image format
for file in files:
    #Convert image to string for a specific language, by default lenaguage is english 
    outText = pytesseract.image_to_string(filePath + file, lang='tha', config='--psm 11')

    #Split in Array of String 
    outTextList = outText.split('\n')

    #extract data from above string and prepare Json file
    #jsondata = { "Vendor" : outTextList[1] } 
 
#Save extracted json data into json file
with open(filePath + fileName + '.json','w') as json_file:
                    json.dump(jsondata, json_file)

4. Config values generally used in Pytesseract

Page segmentation modes:
  0    Orientation and script detection (OSD) only.
  1    Automatic page segmentation with OSD.
  2    Automatic page segmentation, but no OSD, or OCR.
  3    Fully automatic page segmentation, but no OSD. (Default)
  4    Assume a single column of text of variable sizes.
  5    Assume a single uniform block of vertically aligned text.
  6    Assume a single uniform block of text.
  7    Treat the image as a single text line.
  8    Treat the image as a single word.
  9    Treat the image as a single word in a circle.
 10    Treat the image as a single character.
 11    Sparse text. Find as much text as possible in no particular order.
 12    Sparse text with OSD.
 13    Raw line. Treat the image as a single text line,bypassing hacks that are Tesseract-specific.