Numerical Methods for Estimation of Zero Moment Extremal Value Variations

In this work two numerical methods for estimation of zero order moment (number of square grid pixels lying in object image) variations due to image position changes in a raster are described. The first method is based on the construction of a universal-computing network (grid) in the domain of function under investigation and getting guaranteed intervals for its extremal values and variation. In the second method a special computing network is built up taking into account the concrete problem distinctions and providing the necessary estimations with the given level of credit. Taking together, both methods can be considered as the fundamental approaches to the problems of numerical estimation of extremal values and variation of a function resisting complete analytical methods.