Optimal grayscale morphological filters under the LMS algorithm

This paper presents a method for determining optimal grayscale function processing (FP) morphological filters such as close-opening (GO) and open-closing (OC). The proposed method utilizes the matrix representation of both CO and OC with the block basis matrix. With this representation, CO and OC are accomplished by local matrix operations rather than cascade operations. The LMS and back-propagation algorithms are utilized for obtaining the optimal block basis matrix for CO and OC. Some results of optimal morphological filters applied to 2D images are presented.<
>