Implications of a ''Noisy'' observer to data processing techniques

摘要: It is attempted to show how an internal noise source (darklight and threshold jitter) would tend explain experimental data concerning the visual detection of noise-limited signal in diagnostic imaging. The interesting conclusions can be drawn that sets upper limit utility processing techniques designed reduce image noise. Moreover, there should instances where contrast enhancement may far more useful human observer than corresponding reductions amplitude, especially at high count rates (sigma/sub p/ less or equal sigma/sub D/). Then too, limitations imposed on by source, point towards need for additional methods (e.g. computer/microdensitometer) interpreting images photon density so highest possible ratio might obtained.