Structural optimization of pseudogradient algorithms for estimating image parameters
А. Г. Ташлинский · Pattern Recognition and Image Analysis · 2006
The possibility of reducing the computational costs for recursive estimation of the parameters of interframe geometric deformations of images (IGDIs) by means of a structural optimization of algorithms is considered. This optimization is based on finding a subdomain of the domain of definition of the parameters that contains an extremum of a goal function (GF). A situation is analyzed when the domain of definition of the estimated parameters is several times greater than the operating range of the algorithms. Basic relations for the probability of erroneous choice of the subdomain and for discrete probability distributions of the number of iterations of an algorithm are found.