Feature selection algorithm and cobweb correlation

In this paper a method of feature selection is proposed where two different measures of cumulative between-class distance are considered. The algorithm has been derived to select an optimal set of translations for computing of the correlation function between couples of pixels. The texture features obtained from the used model result from the local context evaluation on the selected set of image points. This approach has been inspired with one type of Haralick's parameter. The selection algorithm is addressed, and its good properties are discussed. The results are presented using the images of magnetic domain structures from optical microscopy.