How about this:
#!/bin/bash
# test whether first argument exists and is a regular file, if not exit
[ -f "$1" ] || exit 1
# put 1440 bytes per output file
split -b 1440 "$1" "$1". &&
# merge files again
cat "$1".* >"${1%/*}/new-${1##*/}"
A script begins with a shebang specifying the program it should be run with, in this case that’s /bin/bash. First we test for the first argument to the script to be a regular file and exit if it’s not. The split command splits the file given as the first argument and creates output files like image.jpg.aa for a file image.jpg. Only if split exited successfully, the cat command merges them again using bash’s Pathname Expansion and Parameter Expansion and saves the output (i.e. the merged file) as new-image.jpg for a file image.jpg.
Save this script as e.g. splitmerge.bash, make it executable with chmod +x splitmerge.bash and run it with:
splitmerge.bash image.jpg
All output files are created in the input file’s directory regardless your current directory. If the script and/or the file to split is not located in your current directory, use absolute paths instead, e.g.:
~/scripts/splitmerge.bash /home/thmk/test/image.jpg