[FIXED] Getting text from low-quality gif file with tesseract

Issue

import pytesseract


import requests
from io import BytesIO

from PIL import Image,ImageOps

image_url = 'gif'

def optimize_and_ocr_from_url(image_url, tesseract_config="--psm 6"):
    # Download the image from the URL
    response = requests.get(image_url)
    image = Image.open(BytesIO(response.content))

    # Optimize the image (resize, convert to grayscale, etc.)
    optimized_image = optimize_image(image)

    # Use Tesseract OCR to extract text from the optimized image
    text = pytesseract.image_to_string(optimized_image, config=tesseract_config)

    # Print the OCR result to the console
    print("OCR Result:")
    print(text)

def optimize_image(image, target_resolution=(109, 50)):
    # Resize the image to the target resolution
    resized_image = image.resize(target_resolution)

    # Convert the image to grayscale
    grayscale_image = ImageOps.grayscale(resized_image)

    return grayscale_image

# Example usage
optimize_and_ocr_from_url(image_url)

I am trying to convert this type of images to text This is the type of image i am trying to convert to text

I tried to grayscale,tried to play with the tesseract_config but it doesn't seem to work.There is just empty response.I tried other solutions on stackoverflow but couldnt managed to do it. I use windows

Solution

I tried to use your gif directly with tesseract (instead of pytesseract) in console/terminal/bash (on Linux)

$ tesseract input.gif stdout

Warning: Invalid resolution 0 dpi. Using 70 instead.
Estimating resolution as 270
Empty page!!
Estimating resolution as 270
Empty page!!

Next I tried to convert it with imagemagick and it seems it has few problems:

• it needs to add information about dpi to file but it seems gif can't keep it so I had to convert to png.
• it is too small so I resized it to 200x24
• it is transparent (it doesn't have background) so (to make sure) I addad white background (but it seems tesseract doesn't need it, but it is more readable when I check it in any image viewer)

$ convert input.gif -resize 200x24 -units pixelsperinch -density 72 -background white -flatten output.png

output.png

Now it recognizes text but there is another problem:
There is no top margin so it has problem to recognize 7 - it detects /

+3/0 3/7 312208

So I used -gravity south -extent 200x28 to add margin

$ convert input.gif -resize 200x24 -units pixelsperinch -density 72 -background white -flatten -gravity south -extent 200x28 output-with-margin.png

and now tesseract detects number correctly.

output-with-margin.png (Stackoverflow has white background so you can't see new margin)

---

You can try to do the same with Pillow or you may use subprocess.run() to run directly imagemagick or you may try to use Python module Wand which uses imagemagick

---

EDIT:

It doesn't need top margin if I use digits, + and space

$ tesseract output.png stdout -c tessedit_char_whitelist='+0123456789 '

---

EDIT:

I checked all options in tesseract and it has --dpi - and it can use gif with option --dpi 72 but not with --dpi 70 (maybe if you resize even more then 70 will work).
But image still has to be resized to 200x24, and may need white background to recognize 7.
I didn't test combination: with top margin and without background.

$ tesseract output.gif stdout --dpi 72 -c tessedit_char_whitelist='+0123456789 '

---

EDIT:

In Tesseract documentation you can see Improving the quality of the output | tessdoc

Answered By - furas