Variational algorithms of measurand reconstruction based on entropy-like criteria

The reconstruction of a measurand on the basis of measurement data which are subject to systematic distortions and random errors is a problem often met in instrumental applications. Different regularization techniques are used to deal with its ill-conditioning. In this paper the applicability of variational methods based on entropy-like criteria is studied. The quality of reconstruction is assessed using synthetic measurement data as well as real-world spectrophotometric data. Two proposed algorithms are compared with other methods frequently used for measurand reconstruction