Tesseract over high compressed images

Skip to first unread message

tomeu vidal

Jul 16, 2022, 9:04:57 AMJul 16
to tesseract-ocr

Hi there,

I have about 60000 images (.png and .jpg) highly compressed with Pngquant and Cjpeg
respectively, both agresive lossy image compressors.

I'm optaining very bad results with Tesseract compared with Abbyy finereader.

I'm executing a batch iterative script on windows 10 over the files whit this command:

"C:\Program Files (x86)\Tesseract-OCR\tesseract.exe" "%%F" "z:\%%~nF_T" --oem 1 --psm 12 -c preserve_interword_spaces=1

I'm convinced that there has to be some way for Tesseract to perform better than
finereader. But it's the first time I use it and I don't know how. I attach a sample image.

Any suggestions from you guys?

all the best.
Reply all
Reply to author
0 new messages