FRACTIONAL CENTRAL MOMENT METHOD FOR MOVEMENT-INVARIANT OBJECT CLASSIFICATION

Within the context of moment methods for movement-invariant feature ve ctors the authors derive a new 'low-level' moment method capable of re taining scale and translation properties demonstrated by the alternati ve central moment low-level moment method. The new low-level moment me thod, denoted fractional central moments (FCM), provides a path for ex pressing the high-level moment method of pseudo-Zernike moments in ter ms of low-level moments, thus defining a set of feature vectors provid ing invariance to translation, scale and rotation of objects contained within the image space. The FCM representation provides more moment m ethod terms per order than alternative low-level moment methods, thus it is shown to demonstrate greater image encoding/descriptive properti es at a given maximum moment method order. The authors quantify differ ences between central and fractional central moment methods using disc riminant analysis as applied to a specific data set proposed for the p urpose of investigations described in a sequel paper quantifying neura l network generalisation ability.