Premium
Rotational band identification: A model‐based detection and estimation approach
Author(s) -
Candy J. V.,
Manatt D. R.,
Barnes F. L.,
Becker J. A.,
Henry E. A.,
Brinkman M. J.
Publication year - 1991
Publication title -
international journal of imaging systems and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.359
H-Index - 47
eISSN - 1098-1098
pISSN - 0899-9457
DOI - 10.1002/ima.1850030102
Subject(s) - maxima and minima , metastability , computer science , energy (signal processing) , noise (video) , signal (programming language) , deformation (meteorology) , rotational energy , artificial intelligence , physics , image processing , identification (biology) , algorithm , nuclear structure , biological system , computational physics , statistical physics , image (mathematics) , mathematics , mathematical analysis , atomic physics , quantum mechanics , botany , meteorology , biology , programming language
Nuclear theory predicts that secondary minima develop in the nuclear potential energy surface of some nuclei as they are stretched or elongated. Some minima are deep enough to support quasistationary states. The states are metastable because the decay is inhibited by the required shape change. Detection of these metastable states and their associated rotational bands allows us to investigate nuclei at large deformations and test existing theoretical predictions. The identification of these difficult‐to‐locate structures will permit the detailed study of nuclei at extraordinarily high deformation, providing new insights into nuclear structure. We discuss the application of model‐based image processing techniques to detect and enhance the nuclear rotational bands in noisy experimental measurements. The aim of the experiment is to identify the shape isomer by detecting the presence of the associated rotational band pattern. Unfortunately, the band signal is only ≈1% or less of the amplitude of similar disturbance signals due to the nuclear states of normal deformation; thus the signal‐to‐noise ratio is quite poor. Thus it is necessary to investigate image processing techniques to detect and extract the predicted band pattern. Using a model‐based approach, that is, characterizing a particular band by a simple mathematical model and incorporating it into an image processing scheme, we show that we are able to solve the band pattern detection and enhancement problems from both synthesized and measured experimental data.