CODEWITHSUNDEEP
pythonopencvimage-processingocrpython-tesseract
MarthaF
MarthaF

Reputation: 187

Pytesseract fails to recognize digits from image

Pytesseract fails to recognize digits 6 and 8. It recognizes

The script used:

config= "-c tessedit_char_whitelist=01234567890.:ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz -psm 3 -oem 0"
text = pytesseract.image_to_string(image, config=config)

Tried also using the different psm number from 1-12 but no luck. Increasing contrast results in more numbers not recognized:

kernel = np.ones((2,2),np.uint8)
dilation = cv2.dilate(im, kernel)#,iterations = 1)
text = pytesseract.image_to_string(dilation, config=config)

Raw data:

How the raw data looks like

After running the script:

After running the script

After running new script:

After running new script

Upvotes: 2

Views: 1858

Answers (1)

nathancy
nathancy

Reputation: 46660

Some preprocessing to clean/smooth the image before throwing it into Pytesseract can help. Specifically, morphological operations to close small holes and remove noise can enhance the image. Also applying sharpening filters may help as well. Also adjusting the kernel size or type may help. I believe --psm 6 is the best here since the image is a single uniform block of text. Here's what I get after a simple morph close

enter image description here

import cv2
import pytesseract

pytesseract.pytesseract.tesseract_cmd = r"C:\Program Files\Tesseract-OCR\tesseract.exe"

image = cv2.imread('1.png',0)
thresh = cv2.threshold(image, 150, 255, cv2.THRESH_BINARY_INV)[1]

kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (2,2))
close = cv2.morphologyEx(thresh, cv2.MORPH_CLOSE, kernel)
result = 255 - close

data = pytesseract.image_to_string(result, lang='eng',config='--psm 6')
print(data)

cv2.imshow('thresh', thresh)
cv2.imshow('result', result)
cv2.imshow('close', close)
cv2.waitKey()

Upvotes: 2

Related Questions