Two models for partially coherent imaging.

This paper discusses the physical difference between two image formation approaches for partially coherent imaging. In one approach, the pupil function is shifted according to illumination condition as in the transmission cross coefficient (TCC) approach, whereas in the other approach, the object spectrum is shifted. Although the two approaches result in identical images, they are built on distinct physical models. Eigenfunction analysis reveals that the TCC approach is built on an artificial optical model only for image calculation. Therefore, the two approaches are not interchangeable except for image calculation. Such an example is found in calculating the entropy in an imaging system.