Sparse Signal Representation for a single-image Super Resolution: The research on image Statistics gives a forward step to represent the image patch in a better way

Jadhav, Dr. Jagannath

ISBN 10: 6202803177 / ISBN 13: 9786202803175

Editore: LAP LAMBERT Academic Publishing, 2020

Riguardo questo articolo

New.

Codice articolo ERICA82962028031776

Dati bibliografici

• Titolo: Sparse Signal Representation for a single-image Super Resolution: The research on image Statistics gives a forward step to represent the image patch in a better way
• Editore: LAP LAMBERT Academic Publishing
• Data di pubblicazione: 2020
• Lingua: Inglese
• ISBN 10: 6202803177
• ISBN 13: 9786202803175
• Rilegatura: paperback
• Condizione: New
• Tipo di libro: book

Descrizione articolo

Riassunto

Many effective approaches designed to solve ill-posed and ill-conditioned problem had deficiencies to fulfill the needs of point spread function (PSF), which is hard to get into the practical situation all the time. So this project introduces a method called as Sparse signal representation for a single-image Super Resolution. The research on image Statistics gives a forward step to represent the image patches in a better way, as a sparse linear combination of elements, which are chosen from complete dictionary. From the coefficients of the sparse representation are utilized to construct the high-resolution output image. Here it trains two dictionaries jointly for the low-and high-resolution image patch, which produces two individual dictionary and it shows that the sparse representations for low- and high-resolution is same. To produce a high resolution image patch, the sparse representation can put together two trained dictionaries of the low- and the high-resolution image patch. A large amount of image patch pair are sampled here, by decreasing the computational cost significantly.