Converting would be better since you can add images with composite or layers.
Here is a very crude example, and I'll probably start with a longer section of the wall so you can crop it when resizing. There will still be quite a lot of user input to fix the locations, since I click that you do not always have the same number of walls of the same length. I would write code to enter details into the form, since it would be easier than changing the code every time.
convert k0jbqs.jpg \ ( 313386r.png -thumbnail x25 ) -gravity west -geometry +0+30 -composite \ ( 313386r.png -thumbnail x25 ) -gravity center -geometry +80+30 -composite \ ( 313386r.png -thumbnail x25 ) -gravity east -geometry +0+30 -composite \ output.png
You will need to rip out the lines, I just added them to read the code.
NOTES: Thumbnail resizes images of bricks; you can forget gravity and just use -geometry, and numbers are positions from the upper left corner and -composition puts a new image on top of the previous image.
Bonzo source share