Tytuł pozycji:
The influence of mismatching of input data on FIR filter coefficients generated by the neural network
In this article, the authors have presented the problem of inexact matching of input data in the learning process of a neural network generating a FIR filter in the form of a 5x5 mask. The obtained filter has the task of improving the quality of digital images acquired from acquisition devices, the construction and method of operation of which are based on a low-quality CCD matrix.