Performance analysis of an adaptive flat multi-aperture computational imaging sensor

The traditional single aperture optical sensors are constrained by their large form factor that makes them bulky. This constraint restricts their utility in many scenarios where their pervasive use would otherwise be beneficial. Existing architectures like TOMBO, with multiple apertures and optimal reconstruction algorithms, gives a flat sensor but they are not adaptive to information content in the scene. An adaptive, flat multi-aperture computational imaging sensor architecture named PANOPTES is presented. It utilizes the arrays of micro mirrors to steer the field of view of many low resolution imaging sensors called SIs. This research analyzes the implications of scaling down the optical systems, to short working distances (to achieve flat form factor), on the performance of the system in terms of its sensitivity and the spatial resolution. The radiometric performance analysis and the frequency response analysis are performed in order to determine the SNR and the modulation transfer function (MTF) of the SI, respectively. The effects of using micro-mirror arrays at the sensor pupil are investigated in terms of the degradation in the MTF. The overall performance improvement scheme based on adaptive and redundant resources is proposed. A framework to achieve an all encompassing performance metric for a multi-aperature computational imager is presented.