Sparse representation using compressed sensing via deep learning

D.M. Annie Brighty Christlin, M. Safish Mary


Since the patients are not static during the MRI acquisition, the image formation process can create some artifacts that could reduce the photograph quality. The Compressed Sensing (CS) mechanism is hired for reconstructing the unique picture from the given sparse data. Accordingly, CS can be applied to reduce the acceleration time for an MRI test considering the patient's health. So the sensing process is carried out by way of a projection matrix for reconstructing the sparse signals from a few numbers of measurements. The CS guarantees the healing of an authentic picture with an excessive possibility based totally on random gaussian projection matrices. However, sparse ternaries projections (Latin word)-1, zero, +1 are more apt for hardware implementation. The proposed deep learning technique is hired in this article to acquire very CS sparse ternary projections. The STDAENN architecture incorporates the sensing layer for the projection matrix and a reconstruction layer for non-linear sparse matrix continuously by auto-encoder. The overall performance of the proposed STDAENN method is compared with the present strategies primarily based on the imply top signal-to-noise ratio (PSNR) to check the picture nicely.

Full Text: PDF

Published: 2020-11-16

How to Cite this Article:

D.M. Annie Brighty Christlin, M. Safish Mary, Sparse representation using compressed sensing via deep learning, J. Math. Comput. Sci., 11 (2021), 61-73

Copyright © 2021 D.M. Annie Brighty Christlin, M. Safish Mary. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

J. Math. Comput. Sci.

ISSN: 1927-5307

Editorial Office:


Copyright ©2020 JMCS