Multi-scale Shape Feature Extraction of Erythrocyte SEMed Image Through Curvelet Transform

Abstract

The shape of red blood cell plays an important role to its deformability and filterability. In real clinic blood-related disease diagnosing case, the irregularity and deformation give rise to huge challenges. Thus the shape feature extraction of red blood cell will make sense for medical image compute aided diagnosing. In this paper, the different shape representation was introduced firstly. And then some background knowledge about Curvelet transfrom was reviewed. We conduct wrapping-based curvelet transform on red blood cell image and obtained its reconstruction with transformed multi-scale coefficients to extract the shape feature and draw some conclusion, as well as some future works are proposed.