[FIXED] Tesseract bounding box distance calculation

Issue

I am trying to apply tesseract OCR to extract text from an image. At first, I tried image_to_data, which returned text but in some cases, it gave mispled char. Then I tried image_to_boxes which returned each char very correctly. But my problem is, I need to concatenate those characters to make full words. Can anyone suggest to me any idea how can I do this? I meant to say that I need to make three words from this dictionary such as "~Phone" , ":", "+88-02-5042248-50".

And finally, can anyone please explain what does mean by left, bottom, right, and top in this dictionary? Can I use them for finding distance between two char?

Solution

import cv2
import numpy as np
import pytesseract
# Load image
image = cv2.imread('muTYX.jpg')

pytesseract.pytesseract.tesseract_cmd = r'C:\Program Files\Tesseract-OCR\tesseract.exe'
text = pytesseract.image_to_string(image, lang = 'eng', config='--psm 7 --oem 3 ')

text = (text.split('\n'))
print(text[0].split(' '))

Output: ['Phone', ':', '+88-02-55042248-50']

And the dictionary of image_to_boxes is like:

#(index of character, start x-axis, start y-axis, end x-axis, end y-axis)
P 53 50 64 76 0 
h 63 50 75 77 0
...

There is more info in the following website : https://nanonets.com/blog/ocr-with-tesseract/

Answered By - CodingPeter