Though it is not done from a behavioral viewpoint, the work of Zenon Pylyshyn on visual imagery back in the 1970's (I don't have any specific references) may be helpful in trying to measure that which is termed visual imagery.  In principle it is not directly measurable and will be inferred, but the cognitive psychologists have come up with some ingenious methods of infererence and this work should not be ignored by behaviorists, though we may interpret it differently.

 

Bob Stein