Gu Y-H & Tjahjadi T

IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics, Vol. 32, No. 4, August 2002, 443-454

Abstract

This paper presents a novel method which uses a special type of multiscale isotropic bandpass filters to detect image landmark and corner features. The paper is mainly contributing to the solutions of the following problems: 1) To define a general family of feature detectors under a unified framework that are able to enhance and detect the desired features; 2) To explain theoretically and experimentally why these feature points can be detected by the proposed detectors; 3) To extend the detectors to multiscale versions for jointly achieving good detectability and localization where an automatic scale selection method is applied. The paper then presents several applications for detecting landmarks and corners using the proposed methods, in order to illustrate their usage. These include detecting landmarks from gesture images (of face and hand), from airborne and vehicle-borne IR landmine images, and from images containing object corners. Experiments have been performed using the proposed detectors to these applications. Some comparisons and evaluations have also been performed. The results have demonstrated the effectiveness of the proposed detectors in terms of feature detectability and localization.