Choice of the objective function for pseudogradient measurement of image parameters

Pseudogradient Estimation of Digital Images Interframe Geometrical Deformations

“…We can achieve even more calculations reduction using various interpolations for the prediction. When performing the interpolations at the current iteration of the algorithm the estimates α obtained at the preceding iteration are employed (Minkina et al, 2007). Then, to find the pseudogradient at the t -th iteration of the algorithm it is enough to use a local sample Under these assumptions the pseudogradients obtained on the basis of relations (6) and (7) will become…”

“…G. Tashlinskii (2007). Pseudogradient Estimation of Digital Images Interframe Geometrical Deformations, Vision Systems: Segmentation and Pattern Recognition, Goro Obinata and Ashish Dutta (Ed.…”

“…In this case we can estimate the parameters α on the basis of a given realization Z analysis of the image to be observed by means of some adaptation procedure which minimizes ) , J( ) J( Z α = α for the given realization Z . However, it is reasonable to avoid this intermediate state of the research and estimate α directly on values ) , J( Z α (Polyak & Tsypkin, 1984): images, the analysis of approaches (Tashlinskii, 2000;Minkina et al, 2007) to the synthesis of algorithms of IIGDs estimation in real time showed that it is expedient to seek a decision, satisfying the requirements of simplicity, rapid convergence and efficiency in various real situations, among recurrent non-identification algorithms. The pseudogradient algorithms (PGAs) constitute the most representative class of such algorithms.…”