How to determine compression and gain of local tone mapping

[zhangyulb]

Hi,

I find that the parameters compression and gain of the modified exposure fusion algorithm quite sensitive to the input image's lighting condition. Is there any good idea to adaptively adjust them?