Delaunay triangulation based image enhancement for echocardiography images

A novel image enhancement approach for automatic echocardiography image processing is proposed. The main steps include undecimated wavelet based speckle noise reduction, edge detection, followed by a regional enhancement process that employs Delaunay triangulation based thresholding. The edge detection is performed using a fuzzy logic based center point detection and a subsequent radial search based fuzzy multiscale edge detection. The edges obtained are used as the vertices for Delaunay triangulation for enhancement purposes. This method enhances the heart wall region in the echo image. This technique is applied to both synthetic and real image sets that were obtained from a local hospital.