How are you thinking of using the pictures? Never heard of such a database and frankly doubt it´s worth. What would be "right"? Considering only one individual then an optimal posture should allow free breath with a minimum activation of muscles not necessary for sitting. But comparing one person to another? Different body more likely means a different posture to some degree. A database with weight bearing, on the other hand, may offer comparability.

thank.

Do you think thermal images looking for

muscle activation should work?

I heard of this application. Anyway there are some huge bias:

- ventilation (i.e. high convection) if your subject is moving

- sweathing

you could try to correlate your thermal signal to another signal (e.g.emg).