Improving the quality of medical images in shearlet domain

摘要

In clinical diagnostics, the diagnostic approach based on images obtained from equipment of which medical machine (diagnostic imaging) plays an important role. However, most of medical images have blur combined with noise. There are many reasons to create blur combined with noise in medical images such as the environment, capture device, technician's skills, etc. This problem will affect the process diagnose. In this paper, we proposed a new method to improve the quality of medical images. The proposed method uses cycle spinning combined with Kernels set in shearlet domain. Our algorithm removes the noise and blur details in shearlet domain, and must not the value of point-spread function (PSF). The proposed algorithm not only significantly improves the edge accuracy, but also reduces the loss of information in medical image.