How to set the binary mask in OpenCV python correctly?

Question

I have an image img. I also have a mask with value 255 at all the places where I want to retain the pixel values of img are 0 at all other places.

I want to use these two images viz. the mask and img such that I create a matrix with original img values at places where the mask is 255, and the value -1 at all places where mask is 0.

So, far, I have written this:

maskedImg = cv2.bitwise_and(img, mask)

but the maskedImg has 0 at all the places where the mask has 0. How can I get the value -1 instead of 0 at all the other places using a fast bitwise operation?

swap the mask around (bitwise not) and bitwise or it with the image — Mad Physicist
– Mad Physicist, Commented Dec 17, 2017 at 20:53
@MadPhysicist thanks for the comment. If I swap the mask and the or the mask with the image, I will have the value 255 at all the places outside the mask. I want the value -1 at all those other places outside the mask. — London guy
– London guy, Commented Dec 17, 2017 at 21:05
not sure why you prefer -1... but OpenCV usually uses uint8 for most of the images, including masks, so probably you have to change the type frist.... for it to be able to have -1 — api55
– api55, Commented Dec 17, 2017 at 21:48
Also, as @api55 pointed out, you need to indicate the type of your image and tell us what how far you are willing to go in of conversions. — Mad Physicist
– Mad Physicist, Commented Dec 17, 2017 at 22:04

Kinght 金 · Accepted Answer · 2017-12-18 06:55:27Z

2

I don't know what is your image's dtype. Default is np.uint8, so you cann't set -1 on the result, it underflows to -1 + 256 = 255. That is to say, if the dtype is np.uint8, you cann't set it to negative value.

If you want to set to -1, you should change the dtype.

#masked = cv2.bitwise_and(img, mask).astype(np.int32)
masked = np.int32(img)
masked[mask==0] = -1

edited Dec 18, 2017 at 6:55

answered Dec 18, 2017 at 6:48

Kinght 金

18.4k5 gold badges63 silver badges77 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

How to set the binary mask in OpenCV python correctly?

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related