OpenCV & Python: Cover a colored mask over a image

Question

I want to cover a image with a transparent solid color overlay in the shape of a black-white mask

Currently I'm using the following java code to implement this.

redImg = new Mat(image.size(), image.type(), new Scalar(255, 0, 0));
redImg.copyTo(image, mask);

I'm not familiar with the python api.

So I want to know if there any alternative api in python. Is there any better implementation?

image:

mask:

what i want:

Bereniceberenson

Ok, Can you also show what have you tried so far ? – Andra

Have you tried some permutations of blending the two images? :) – Nephritis

Answer 1

Now after I deal with all this Python, OpenCV, Numpy thing for a while, I find out it's quite simple to implement this with code:

image[mask] = (0, 0, 255)

-------------- the original answer --------------

I solved this by the following code:

redImg = np.zeros(image.shape, image.dtype)
redImg[:,:] = (0, 0, 255)
redMask = cv2.bitwise_and(redImg, redImg, mask=mask)
cv2.addWeighted(redMask, 1, image, 1, 0, image)

Bereniceberenson

Regarding your updated answer, I think you still need addWeighted otherwise you won't overlay with transparency. – Wardmote

Answer 2

The idea is to convert the mask to a binary format where pixels are either 0 (black) or 255 (white). White pixels represent sections that are kept while black sections are thrown away. Then set all white pixels on the mask to your desired BGR color.

Input image and mask

Result

Code

import cv2

image = cv2.imread('1.jpg')
mask = cv2.imread('mask.jpg', 0)
mask = cv2.threshold(mask, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)[1]
image[mask==255] = (36,255,12)

cv2.imshow('image', image)
cv2.imshow('mask', mask)
cv2.waitKey()

Sesquiplane

Answer 3

this is what worked for me:

red = np.ones(mask.shape)
red = red*255
img[:,:,0][mask>0] = red[mask>0]

so I made a 2d array with solid 255 values and replaced it with my image's red band in pixels where the mask is not zero. redmask

Kielce