Performance measures for 3d shape similarity

Question

I appreciate please if I got help with this:

my database is indexed like this:

[0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3