The Concept of Superresolution

Abstract

Resolution is characterized by the finest detail that can pass through the system without being distorted. The motivation of the superresolution field is to handle the nonresolved details using given a priori information about the input signal [36]. Such an observation is related to the diffraction limitations of the medium. A more correct point of view is to examine the meaning of the term “system.” Its definition should be the medium that connects the optical signal (let’s say optical) to the electronic output signal coming from the detector. Thus, the term “system” may be subdivided into three subgroups. The first group is the free space medium in which the optical signal propagates. The resolution obtained in this subgroup is the diffraction resolution [36]. Then this signal is captured by the CCD camera. The spatial information is again distorted since the pixels of the CCD have a finite and nonzero size. This type of superresolution is termed geometrical resolution [37]. After the capture of the optical signal it is converted into an electronic signal. Here, the quality of the detector comes into play: its sensitivity, its dynamic range, its shot-noise level, its readout noise, and its noises which are related to generation/recombination processes in the detector. All these noises developed in each cell of the detector again damage the obtained output signal and we call them the noise equivalent resolution. The resolution improvement of the final electronic readout signal is a combination of improvements achieved in each one of the three stages. The process of improvement is the real meaning of the term superresolution.KeywordsOptical SignalPolarization CodificationReadout NoiseGeometrical ResolutionDiffraction ResolutionThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.