Calibrated Digital Imaging Systems

The concept of simulating chosen information channels of a linear incoherent imaging system by projecting reference signals is proposed and the system's ability to restore severely distorted but stationary images is experimentally demonstrated. A particular instrument's object field is divided into N discrete 'points' whose individual simulated outputs is measured at N particular sampling points; the NxN output matrix F is the transfer matrix of the system, and being empirically derived, it precisely accounts for all diffractions, aberrations, distortions, peculiarities of the chosen object points and sensor size and distribution. As such F can become the basis for eliminating algorithmhardware mismatch in various restoration schemes. Calibration allows treating the instrument as an unknown 'black box', and imaging through a severely and arbitrarily distorting hardware is demonstrated. Other uses of F are to add or subtract the outputs of separate instruments viewing the same object on a pixel-by-pixel basis, and as a graphically or digitally presented measure of the instrument's local transfer function, and hence of image quality.