Optimizing filter banks for supervised texture recognition

Abstract

Two criteria for invariant supervised texture segmentation based on multi-channel approaches are introduced. The texture segmentation is carried out by feature extraction using multi-channel Gabor filtering and classification with symmetric phase-only matched filtering. For the feature extraction highly efficient filter banks are required that enable clear distinction between feature vectors representing different textures in order to achieve a high classification rate. For the design of the filter banks, the variances of the frequency components must be maximized. The spar hyper volume spanned by the normalized feature vectors representing different textures must be maximized as well. These two criteria provide guidelines for filter bank design.