BEMD-Unsharp Masking for retinal angiography image sharpening

Abstract

Image sharpening is the essential preprocessing step when improving the angiographies retinal image quality. It is helpful for the vessel retinal analysis and for improving the quality of their. For this reason, we propose a new technique for image sharpening based on Unsharp Masking (UM), and Bidimensional Empirical Mode Decomposition (BEMD). Firstly, the image is decomposed into a set of bidimensional intrinsic mode functions (BIMFs) and the residual image. Afterward, a weighting mask is achieved from an edge map multiplied by a compensation factor. Then, we apply the weighting mask to the first mode. Finally, we perform the reconstruction of the sharpened image by summing the compensated BIMF1 with the rest of the other modes and the residual image. The proposed scheme is enhanced by means of deringing's step to overcome the overshooting introduced during image sharpening. The obtained results proved that the proposed approach is effective to sharpen retinal images.