Hi,
I believe there's no need to use composite. In fact I would first extract the needed geometry information (for both images) by using identify:
Code:
identify -format "%w" image.png ## gives the width
identify -format "%h" image.png ## gives the height
Then I would use the option -frame with convert, in order to selectively add the needed frame. The option -mattecolor will specify the color of the frame itself:
Code:
convert IN.png -mattecolor white -frame "${width}"x"${height}" OUT.png
Note that width and height parameters of the option -frame define the size of the frame alone, not of the output image. For instance, in order to get your IN.png image from 1000x1200 to 1600x1200 you will need:
Code:
convert IN.png -mattecolor white -frame 300x0 OUT.png
Basically I would use the extracted geometry information to calculate the needed frame, and put everything in a nice script (ImageMagick rocks).
Best regards,
Philip