You can only estimate a stop difference reliably if the scene illumination changed mostly as a uniform brightness scaling. In that case, compare a representative luminance value from both images and convert the ratio to stops:

stops = log2(L2 / L1)

where L1 and L2 are luminance measurements from corresponding image content. If one frame is twice as bright, that is +1 stop; 4× is +2 stops; 1/2 is −1 stop.

However, there is no exact single stop value if the ambient light changed in a non-uniform way. Real lighting changes can alter different parts of the scene differently and can also change color, so one global scalar exposure correction may not fully match the images.

For film, a densitometer can measure density differences, often expressed in base-10 log units; about 0.30 density difference corresponds to 1 stop, though film gamma also affects interpretation.

So: use a luminance ratio only when the lighting change is effectively uniform, and expect it to fail when the light itself changes character across the scene.