A Vector-based Filtering Algorithm for Microarray Image

This paper presents a simple but effective method capable of processing microarray images. By applying nine directional templates to both of the two components of microarray images, the method is able to remove noise impairments while preserving low-contrast signals. This is based on the knowledge that signal spots should have the same edge features in both two components of microarray images, which can be detected by the directional templates, while noise and uneven background do not have this characteristic. The mathematical morphological filtering is then adopted by using a smaller structural element to remove dispersed noise points. The experimental result indicates that the proposed approach yields good performance in removing noise and enhancing spot location.