Combining 2 images of different size

I am trying to combine 2 image of different size. The smaller image is to be fully overlaying a small part of the bigger image without transparency. Like a sticker over a image. How do I go about doing it? I have the specific coordinates of where the smaller image should be overlaid on.

Answer 1

[Walter Roberson]

CompositeImage = BiggerImage;
CompositeImage(FirstRow : FirstRow + size(SmallerImage,1) - 1, FirstColumn : FirstColumn + size(SmallerImage,2) - 1) = SmallerImage;

[Edmund Lua]

I tried your code and got the error "Assignment has more non-singleton rhs dimensions than non-singleton subscripts". Any idea why?

[Walter Roberson]

CompositeImage(FirstRow : FirstRow + size(SmallerImage,1) - 1, FirstColumn : FirstColumn + size(SmallerImage,2) - 1, :) = SmallerImage;

[Edmund Lua]

That solved the problem and I can see it's at the correct position now. However, it's showing just black instead of what the smaller image is supposed to be.

[Walter Roberson]

Please show

size(BiggerImage)
size(SmallerImage)
class(BiggerImage)
class(SmallerImage)

[Edmund Lua]

Here you go

size(BiggerImage)

ans =

     701   574     3

size(SmallerImage)

ans =

      40    66     3

class(BiggerImage)

ans =

uint8

class(SmallerImage)

ans =

uint8

[Walter Roberson]

Seems to work fine in my test.

BiggerImage = uint8(rand(701,574,3) * 255);
SmallerImage = uint8(207 * ones(40,66,3));
FirstRow = 85; FirstColumn=417;
CompositeImage = BiggerImage;
CompositeImage(FirstRow : FirstRow + size(SmallerImage,1) - 1, FirstColumn : FirstColumn + size(SmallerImage,2) - 1,:) = SmallerImage;
image(CompositeImage)

For your data, what shows up for min(SmallerImage(:)) and max(SmallerImage(:)) ?

[Edmund Lua]

I got 0 and 1.

min(SmallerImage(:))

ans =

      0

max(SmallerImage(:))

ans =

      1

[Walter Roberson]

The level 1 in a uint8 image is practically black. You would have a hard time telling the 0 (true black) and the 1 (nearly black) apart in uint8 .

If your original data is in the range 0 to 1 but is double precision instead of uint8, then you should be using

SmallerImage = im2uint8(SmallerImageThatIsDouble);

before starting the above. Or if you do not have the Image Processing Toolbox then

SmallerImage = uint8(255 * SmallerImageThatIsDouble);

[Edmund Lua]

Yes that worked! Thank you so much kind sir.