A FAST DIRECTIONAL FILTER SET FOR PYRAMIDAL IMAGE-BASED ROBOTIC SYSTEMS

In seeking to develop a real-time, pyramidal imaging system for tracki ng and robotic visual servoing applications, inadequacies in the stand ard, commonly a,available kernels for directional filtering have led t o the implementation of an alternative set of simple filters that comb ine to act as oriented edge-contrast operators fulfilling the familiar requirements for both speed and a highly localized response pattern. Although numerous edge operators, kernels and directional filters have been proposed, and low-level filtering is a well-established and matu re area of computer vision, within the application described here, thi s operator set is proving superior to more established methods in effi ciently preserving the response localization requirement while propaga ting the oriented gradient energies through multichannel resolution py ramids for further recursive processing. This paper provides a detaile d discussion of the technique, the context of its development, and sup porting empirical tests and statistical results. The evidence presente d shows that the scheme provides a unified response to both step and l ine edge types, with acceptable performance in the presence of anticip ated noise levels. The approach is pragmatic, of low complexity, and o f general utility. (C) 1998 Published by Elsevier Science Ltd. All rig hts reserved.