Optical design of a coded aperture infrared imaging system with resolution below the pixel limit

Adaptive coded aperture imaging systems can resolve objects that are smaller than the pixel-limited resolution of the detector focal plane array. This is done by combining multiple frames of data, where different frames are taken with different coding patterns on the coded-aperture mask. In the mid-wave infrared the required signal to noise ratio necessitates some form of light concentration. Optical design software has been used to model candidate optical systems with the aim of achieving up to four times resolution enhancement along each linear dimension. As in some other computational imaging systems, the requirements on the optical system are found to be different to those that are normally used in more classical optical designs. The basic needs are a point-spread function of suitable extent that changes gradually with angle and does not vary significantly with the expected changes in input spectra or system temperature. Novel metrics have been derived and used to inform the optical design. The modeling and design trade-offs and resulting performance are discussed.