A Mathematical Operator for Automatic and Real Time Analysis of Sequences of Vascular Images

Abstract

Due to the absolute value involved, the first absolute central moment can be divided into two complementary filters: a positive deviation e p and a negative deviation e n . Both e p and e n can be used separately to highlight edges, lines, line endings, corners and junctions in images. Furthermore, the recovered edge information can be usefully combined to obtain additional information that would not be obtained by varying the parameters of the original filter. The mass center of the first absolute central moment can be also defined and an iterative localization procedure can be developed by exploiting its properties. Mathematical operators derived from the first absolute central moment were used on a video processing device based on a DSP board and they proved to be robust and suitable for real-time implementations.Keywordsedge detectioncontour trackingultrasound image sequences