Investigation of the applicability of Zernike moments to the classification of SDS 2D-PAGE maps

The aim of this work is to investigate theperformance of multivariate classification techniques likepartial least squares–discriminant analysis (PLS-DA) andlinear discriminant analysis (LDA) when using Zernikemoments as global image descriptors, in the classificationof sodium dodecyl sulphate (SDS) two-dimensionalpolyacrylamide gel electrophoresis (2D-PAGE) mapsaffected by different levels of deformation. Synthetic setsof images simulating real SDS 2D-PAGE maps wereanalysed in controlled conditions to obtain information onthe robustness and limits of applicability of the classificationtechniques operating on the basis of a given imagedecomposition method.