How to read black text on black background image through tesseract OCR?

Question

I have black text on black background image and I want to read it through OCR. Unfortunately, OCR can not read it perfectly. The image look like this. I want to convert RGBA value that less than (90, 90, 90, 255) to (255, 255, 255, 255) so it turn B & W. What's the code to convert it?

score 6 · Accepted Answer · answered May 31 '19 at 02:35

What you need to do is make the whole image black and white before letting tesseract do its job.

Read image

import cv2
im_gray = cv2.imread('your_image_here', cv2.IMREAD_GRAYSCALE)

Make it grayscale

(thresh, im_bw) = cv2.threshold(im_gray, 128, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)

"which determines the threshold automatically from the image using Otsu's method, or if you already know the threshold you can use:"

thresh = 127
im_bw = cv2.threshold(im_gray, thresh, 255, cv2.THRESH_BINARY)[1]

Write to disk

cv2.imwrite('bw_image.png', im_bw)

Taken from here

When you use an THRESH_OTSU, automatically THRESH is automatically considered zero. `cv2.threshold(im_gray, THRESH=0, 255, cv2.THRESH_OTSU)` — Milad Shaker, Dec 19 '21 at 19:09

score 0 · Answer 2 · answered May 31 '19 at 02:43

You can transform your gray pixels in white pixels with a simple transformation. If you don't want to use open cv and your image is one channel (gray scale) numpy array:

threshold = 60 # try something between 30 and 150
vect_func = np.vectorize(lambda x: 0 if x == threshold else 255)
black_white_img = vect_func(gray_scale_image)

How to read black text on black background image through tesseract OCR?

2 Answers2