Search code examples
pythonocrtesseractpython-tesseract

Tesseract OCR fails to recognize number 7

I'm trying to analyze a png image containing two digits. Tesseract seems to have an issue detecting the number 7 (seven) in this image. It recognizes it as a 1 (one).

my config is:

config='--oem 1 --psm 7 tessedit_char_whitelist=0123456789'

tesseract --version returns the following

tesseract 4.0.0
leptonica-1.76.0
libgif 5.1.4 : libjpeg 6b (libjpeg-turbo 1.5.2) : libpng 1.6.36 : libtiff 4.1.0 : zlib 1.2.11 : libwebp 0.6.1 : libopenjp2 2.3.0

I've played with the oem and psm values with no change in result. Any help would be much appreciated.

tesseract not recognizing the number 7 (seven) digit

Solution

  • The characters in the image are too big, I guess tesseract was not trained for that.

    Resizing the image works fine

    tesseract cnLRQ.png stdout --oem 1 --psm 7 --dpi 70 digits
18
convert -resize 10% cnLRQ.png cnLRQ-small.png
tesseract cnLRQ-small.png stdout --oem 1 --psm 7 --dpi 70 digits
78