Current research activities in the field of deinterlacing include the selection of suitable deinterlacing methods and the estimation of the exact value of a missing line. This paper proposes a spatio-temporal domain fuzzy rough sets rule for selecting a deinterlacing method that is suitable for regions with high motion or frequent scene changes. The proposed algorithm consists of two parts. The first part is fuzzy rule-based edge-direction detection with an edge preserving part that utilizes fuzzy theory to find the most accurate edge direction and interpolates the missing pixels. Using the introduced gradients in the interpolation, the vertical resolution in the deinterlaced image is subjectively concealed. The second part of the proposed algorithm is a rough sets-assisted optimization which selects the most suitable of five different deinterlacing methods and successively builds approximations of the deinterlaced sequence. Moreover, this approach employs a size reduction of the database system, keeping only the information essential for the process. The proposed algorithm is intended not only to be fast, but also to reduce deinterlacing artifacts.
Fuzzy rough sets hybrid scheme for motion and scene complexity adaptive deinterlacing / G. Jeon, M. Anisetti, D. Kim, V. Bellandi, E. Damiani, J. Jeong. - In: IMAGE AND VISION COMPUTING. - ISSN 0262-8856. - 27:4(2009 Mar), pp. 425-436.
Fuzzy rough sets hybrid scheme for motion and scene complexity adaptive deinterlacing
M. AnisettiSecondo
;V. Bellandi;E. DamianiPenultimo
;
2009
Abstract
Current research activities in the field of deinterlacing include the selection of suitable deinterlacing methods and the estimation of the exact value of a missing line. This paper proposes a spatio-temporal domain fuzzy rough sets rule for selecting a deinterlacing method that is suitable for regions with high motion or frequent scene changes. The proposed algorithm consists of two parts. The first part is fuzzy rule-based edge-direction detection with an edge preserving part that utilizes fuzzy theory to find the most accurate edge direction and interpolates the missing pixels. Using the introduced gradients in the interpolation, the vertical resolution in the deinterlaced image is subjectively concealed. The second part of the proposed algorithm is a rough sets-assisted optimization which selects the most suitable of five different deinterlacing methods and successively builds approximations of the deinterlaced sequence. Moreover, this approach employs a size reduction of the database system, keeping only the information essential for the process. The proposed algorithm is intended not only to be fast, but also to reduce deinterlacing artifacts.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.