A neural system prototype for data processing using modified conjugate directional filtering (MCDF)

Abstract

Modified conjugate directional filtering (MCDF) is a method proposed by us recently for digital data and image processing. By using MCDF, directional-filtered results in conjugate directions can be not only merged into one image that shows the maximum linear features in the two conjugate directions, but also further manipulated by a number of predefined generic MCDF operations for different purposes. In this paper, we report the progressive result of our MCDF study, which shows that a neural system can be used to implement the MCDF operations. We provide the trial work on applying this MCDF neural system to the processing of a digital terrain model (DTM) in central Australia. The trial work shows that the MCDF provides the power of integrating information in different forms, and thus presents more information in a single image than the conventional methods do. © Springer-Verlag 2003.