Abstract

The estimation of a signal from quantum-limited measurements of it is a problem arising in diverse scientific fields, including radionuclide imaging, electron-microscopic autoradiography, gamma-ray astronomy, optical communication, and low light-level photography. As the measurements become more quantum limited, it becomes increasingly important to use accurate mathematical models for the signal, noise, and measurement instrumentation and to use estimation approaches that are dictated by such models. In this paper, we describe a general random point­process model which accurately represents the characteristics of quantum-limited measurements in many situations. We then derive a method for realizing desired estimates within this model.

© 1986 Optical Society of America