Towards a theory for video coding using distributed compression principles