This forum is disabled, please visit https://forum.opencv.org
 | 1 | initial version |
I've two ideas, hope they can help you. 1. Use houghline to find the three lines, and then draw white bars to do the cv2.bitwise_or() with the original image to do away with the three borders. 2. Since there are always 3 borders, you can vertically search the image, and count the pixels line by line. Once current line has black pixels, you get the the vertical axis Y, and then draw the white bar as in idea1 for a accent linewidth to cover the horizontal line. And then do the same action to the two vertical borders. I hope these can help you.
