Information transfer in reconstruction of data from sampled points

Abstract

Evaluation of the fidelity of data, reconstructed from sampled points, particularly in digital form, appears to be still a crucial problem in photogrammetry and cartography. Of special interest is the coherent point sampling because of its suitability for automation. For the analysis of the information transfer in the sampling and reconstruction process an analytical approach has been devised. The concept of the analytical approach is based on the fact that input data can be represented by the Fourier series. The effect of sampling on the fidelity of reconstructed data can therefore be studied on sine waves. Typical for this approach is that the effects of sampling density and measuring errors can be evaluated separately. These two partial effects can be combined by mutual multiplication of the corresponding transfer ratios. For the purpose of planning the procedure is reversal of that applied in evaluation. The specifications for the output should be defined in suitable quantitative terms. The required density of sampling and measuring accuracy can then be established theoretically.