Generalized Feature Extraction Using Expansion Matching 1 Edics Categories: 1.1.5 Image Processing -segmentation 1.1.9 Image Processing -image Analysis

A novel generalized feature extraction method based on the Expansion Matching (EXM) method and on the Karhunen-Loeve transform (KLT) is presented. The method provides an e cient way to locate complex features of interest like corners and junctions with reduced number of ltering operations. The EXM method is used to design optimal detectors for a set of model elementary features. The Karhunen-Loeve (KL) representation of these model EXM detectors is used to lter the image and detect candidate interest points from the energy peaks of the eigen coe cients. The KL coe cients at these candidate points are then used to e ciently reconstruct the response and di erentiate real junctions and corners from arbitrary features in the image. The method is robust to additive noise and is able to successfully extract, classify and nd the myriad compositions of corner and junction features formed by combinations of two or more edges or lines. This method di ers from previous works in several aspects. First, it treats the features not as distinct entities, but as combinations of elementary features. Second, it employs an optimal set of elementary feature detectors based on the Expansion Matching approach. Third, the method incorporates a signi cant reduction in computational complexity by representing a large set of EXM lters by a relatively small number of eigen lters derived by the KL transform of the basic EXM lter set. This is a novel application of the KL transform which is usually employed to represent signals and not impulse responses as in our present work.