Read text below barcode pytesseract python

Question

I am trying to get the number below a barcode in an image. I have tried the same code with some other images and works fine but not for that image Here's the image

And here is the code till now

def readNumber():
    image = cv2.imread(sTemp)
    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
    blur = cv2.GaussianBlur(gray, (3,3), 0)
    thresh = cv2.threshold(blur, 0, 255, cv2.THRESH_BINARY_INV + cv2.THRESH_OTSU)[1]
    kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (3,3))
    opening = cv2.morphologyEx(thresh, cv2.MORPH_OPEN, kernel, iterations=1)
    invert = 255 - opening
    data = pytesseract.image_to_string(invert, lang='eng', config='--psm 6 -c tessedit_char_whitelist=0123456789')
    print(data)
    try:
        data  = re.findall('(\d{9})\D', data)[0]
    except:
        data = ''
    return data

And I used it using this line

readNumber()

Here's another example

This is the last example I promise

I tried this with the third example and it works

img = cv2.imread("thisimage.png")
blur = cv2.GaussianBlur(img, (3,3), 0)
#gry = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
txt = pytesseract.image_to_string(blur)
print(txt)

But how I adopt all the cases to work with the three cases? I tried such a code but couldn't implement the thrid case

import pytesseract, cv2, re

def readNumber(img):
    img = cv2.imread(img)
    gry = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    
    try:
        txt = pytesseract.image_to_string(gry)
        #txt  = re.findall('(\d{9})\D', txt)[0]
    except:
        thr = cv2.adaptiveThreshold(gry, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY, 51, 4)
        txt = pytesseract.image_to_string(thr, config="digits")
        #txt  = re.findall('(\d{9})\D', txt)[0]

    return txt

# M5Pr5         191876320
# RWgrP         202131290
# 6pVH4         193832560
print(readNumber('M5Pr5.png'))

Answer 1

You don't need any preprocessing methods or configuration for the input image. Since there is no artifacts in the image.

import cv2
import pytesseract

img = cv2.imread("RWgrP.png")
gry = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
txt = pytesseract.image_to_string(gry)
print(txt)

Result:

202131290

My pytesseract version is 4.1.1

Update-1

---

The second image requires preprocessing

If you apply adaptive-thresholding:

But the output also consists of unwanted characters. Therefore if you set the configuration to digits, the result will be:

193832560

Update-2

---

For the third image, you need to change the adaptive method, using ADAPTIVE_THRESH_MEAN_C will result in:

191876320

The rest are same.

Code:

import cv2
import pytesseract

img = cv2.imread("6pVH4.png")
gry = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
thr = cv2.adaptiveThreshold(gry, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY, 51, 4)
txt = pytesseract.image_to_string(thr, config="digits")
print(txt)
cv2.imshow("thr", thr)
cv2.waitKey(0)