This experiment illustrates several ways of blending two images. First an average of two images is built. Next, the weighting function applied to combine the pixel is non-uniform, using a gaussian function. Finally, a combination choosing half of the pixels from one image and another half from the other image in a chessboard fashion is described.
Below are the two images to be combined.
![]() |
![]() |
---|---|
a) | b) |
First, applying an average of both images: result = 0.5 * A + 0.5 B
![]() |
---|
Using a gaussian weighting function to blend the two images: result = gaussian * A + (1 - gaussian) * B
![]() |
![]() |
---|---|
a) | b) |
Finally, the weighting image is a binary chessboard image, the same one used in the lesson
Patterns .
![]() |
![]() |
---|---|
a) | b) |
The result of this geometric combination can be seen below. This method has an advantage to display exactly the original pixels of the images. It is your eyes that combine the images together.
![]() |
---|