Tesseract OCR Stopped Working [May 2024]

[digijaco]

Tesseract stopped working in Streamlit. I did some troubleshooting and found out that it does work when running the app locally, but as soon as I deploy the app, the script in bold, below, no longer works. I saw that other users raised the same issue. It stopped working from March 2024.

pytesseract.pytesseract.tesseract_cmd = ‘C:/Program Files (x86)/Tesseract-OCR/tesseract.exe’
response = requests.get(imagetoread)
img = Image.open(io.BytesIO(response.content))
text_img = pytesseract.image_to_string(img)

Can someone please advise?

[Franky1]

@digijaco

pytesseract.pytesseract.tesseract_cmd = ‘C:/Program Files (x86)/Tesseract-OCR/tesseract.exe’

You cannot use any windows paths or executables on streamlit cloud, because it is a linux debian based environment.

[digijaco]

Thank you. What path shall I use (I’m a newbie )? I used this path before, and it’s worked fine for months.

[Franky1]

pytesseract.pytesseract.TesseractNotFoundError - image_to_string

Please share a link to your public github repo. Either the tesseract binary is not installed or could not be found by pytesseract. If tesseract binary is installed and it still does not work, try this code snippet: import pytesseract import shutil import streamlit as st pytesseract.pytesseract.tesseract_cmd = None # search for tesseract binary in path @st.cache_resource def find_tesseract_binary() -> str: return shutil.which("tesseract") # set tesseract binary path pytesseract.pytesser…