0

I am was trying to crop minAreaRect and I got the following result, with black padding to cover up missing parts. I want to replace that black padding with white pixels.

Here is the Image

Dhiraj Kadam
  • 13
  • 1
  • 3

1 Answers1

0

I am replacing the black pixels with white pixels using cv2. I am reading the image in black and white.

img = cv2.imread('input_image',0)

As the img now is a numpy array, The fastest and most concise way to replace black pixels to white is to use Numpy's builtin indexing. Here I am replacing all 0 values in the image to 255(white). The time taken for this operation is 0.00011920928955078125 seconds.

img[img == 0] = 255

This is the resultant image on doing these operations

Amog Chandrashekar
  • 277
  • 2
  • 8
  • That's great, but there's still a black line between the image and the padded white pixels. How can we get rid of them? – Dhiraj Kadam Oct 01 '18 at 10:28
  • @DhirajKadam Those lines are because the pixels in those spots are more than 0 and less than 100, if you want to remove all those black points, change condition to img[img<100]=255, but this adds some amount of distortion to the image(removes some black spots from the image). Please up vote the answer, if you found it to be useful. – Amog Chandrashekar Oct 01 '18 at 10:56
  • Thank you. I'm new so StackOverflow doesn't allow me to upvote yet. However, I accepted it as the answer. – Dhiraj Kadam Oct 01 '18 at 11:03