The colours of the outgoing cables don't tell us anything; they are the usual harmonised phase colours (black, brown, grey) for one winding group, and three other random colours for the other group. You can remove these for now as they will be in the way, but don't disturb the internal connections from the windings.
I hadn't been able to see that there were 9 winding ends in the terminal box. If the two leads on each of UA, VA, WA are the corners of the delta (a.k.a. U1, V1, W1) then those simply need making up into different pairs, and the 230V reconfiguration is complete without having to open the motor. If they are the edges (a.k.a. U2, V2, W2) and the single leads on UB, VB, WB are internally jointed to the corners, we're no further forward and need to get at those internal joints. So it looks like we have a 50% chance that it's a doddle. Now I have to think of a simple test to identify them with minimal equipment and no 3-phase mains.
Referring back to your previous post, no, there's no physical manifestation of star or delta layout in the windings. Those shapes refer to the appearance of the theoretical circuit diagram when particular combinations of winding ends are connected together.
Darkwood makes a good point; not all motors are happy with VFD output waveforms. The short risetimes stress the insulation and older motors sometimes breaks down as a result. I've never actually had this problem myself yet, so I tend to overlook it.