This work is devoted to comparative experimental analysis of different stochastic optimization algorithms for image registration in spatial domain: stochastic gradient descent, Momentum, Nesterov momentum, Adagrad, RMSprop, Adam. Correlation coefficient is considered as the objective function. Experiments are performed on synthetic data generated via wave model with different noise-to-signal ratio.
An optimization criterion is suggested for the plan of counts in a local sample used to determine the pseudogradient of the objective function of estimation quality. The use of the criterion is considered in the case when the object functions are defined as the interframe difference mean square, the covariance, and the interframe correlation coefficient. The optimization is directed at increasing the convergence rate of esti mates of the parameters of geometric interframe image deformations.
A recursive algorithm for detection of radio pulses in real time based on the joint processing of unfiltered signals from spatially distributed receivers, e.g. elements of an antenna array, is proposed. The algorithm is based on a stochastic gradient ascent algorithm estimating the parameters of mutual mismatches between signals received by spatially distributed sensors.
Stabilization of the estimates of signals' time alignment is chosen as one of the detection criteria. Another criterion is based on the analysis of the correlation coefficient between the aligned signals. Signal alignment is performed recurrently in real time. It is shown that if correlation coefficient is chosen as the objective functionshowing the alignment quality we need to estimate only time shift between the received signals. Experiments show the efficiency and the high probability of correct detection of radio pulses location of the proposed algorithms. To increase detection reliability the algorithms can easily be combined.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.