An Algorithm for Real-Time Optimal Photocurrent Estimation including Transient Detection for Resource-Constrained Imaging Applications

摘要

Mega-pixel charge-integrating detectors are common in near-IR imagingapplications. Optimal signal-to-noise ratio estimates of the photocurrents,which are particularly important in the low-signal regime, are produced byfitting linear models to sequential reads of the charge on the detector.Algorithms that solve this problem have a long history, but can becomputationally intensive. Furthermore, the cosmic ray background isappreciable for these detectors in Earth orbit, particularly above the Earth'smagnetic poles and the South Atlantic Anomaly, and on-board reduction routinesmust be capable of flagging affected pixels. In this paper we present analgorithm that generates optimal photocurrent estimates and flags randomtransient charge generation from cosmic rays, and is specifically designed tofit on a computationally restricted platform. We take as a case study theSpectro-Photometer for the History of the Universe, Epoch of Reionization, andIces Explorer (SPHEREx), a NASA Small Explorer astrophysics experiment concept,and show that the algorithm can easily fit in the resource-constrainedenvironment of such a restricted platform. Detailed simulations of the inputastrophysical signals and detector array performance are used to characterizethe fitting routines in the presence of complex noise properties and chargetransients. We use both Hubble Space Telescope Wide Field Camera-3 andWide-field Infrared Survey Explorer to develop an empirical understanding ofthe susceptibility of near-IR detectors in low earth orbit and build a modelfor realistic cosmic ray energy spectra and rates. We show that our algorithmgenerates an unbiased estimate of the true photocurrent that is identical tothat from a standard line fitting package, and characterize the rate, energy,and timing of both detected and undetected transient events.