Analysis of Interest Points of Curvelet Coefficients Contributions of Microscopic Images and Improvement of Edges

Abstract

This paper focuses on improved edge model based on Curvelet coefficients analysis. Curvelet transform is a powerful tool for multiresolution representation of object with anisotropic edge. Curvelet coefficients contributions have been analyzed using Scale Invariant Feature Transform (SIFT), commonly used to study local structure in images. The permutation of Curvelet coefficients from original image and edges image obtained from gradient operator is used to improve original edges. Experimental results show that this method brings out details on edges when the decomposition scale increases.