I am trying to parse text from an image of graded comic books, tesseract seems to be able to parse the smaller text on the image somewhat, but not the largest text on any image i send it.
Here is the code for converting the image. It basically just makes it into black and white.
gray_image = cv.cvtColor(image, cv.COLOR_BGR2GRAY)
threshold_img = cv.threshold(cv.bilateralFilter(
threshold_img, 5, 75, 75), 0, 255, cv.THRESH_BINARY + cv.THRESH_OTSU)[1]
# 6,11,12 perform well, i have tried every psm config
details = pytesseract.image_to_data(
threshold_img, output_type=Output.DICT, lang='eng', config="--psm 6")
I have been able to successfully parse the text when I crop the image to be JUST the grade "label".
One important note is that the grading label can be of different sizes and text can be gray or gold too.
example image, label that needs to be processed is the "8.0" in the top left
If pytesseract would not work in this case, please suggest another image processing tool that would work better in this case! Thanks
