Structure and gradient‐based industrial interpolation using computational intelligence. (17th July 2019)

• Record Type: Journal Article
• Authors: Jeon, Gwanggil
• Other Names: Jeon Gwanggil guestEditor.; Bellandi Valerio guestEditor.; Bakhouya Mohamed guestEditor.; Zbakh Mostapha guestEditor.; Essaaidi Mohamed guestEditor.; Manneback Pierre guestEditor.
• Is Part Of: Concurrency and computation. Volume 32:Number 17(2020)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2019-07-17
• Subjects: area characteristic -- distance -- interpolation -- likeness -- standard deviation; Parallel processing (Electronic computers) -- Periodicals; Parallel computers -- Periodicals; 004.35
• DOI: 10.1002/cpe.5439
• Languages: English
• ISSNs: 1532-0626

Abstract

Summary: Industry and medical imaging are technologies and processes of managing subjective representations of the industry product, interior of a body for analyzing. The deinterlacing is the procedure of transforming interlaced signal into progressive one. In this paper, I propose a new single field deinterlacing method which has three sub‐methods: bilinear method, small filter‐based method, and large filter‐based method. Pixels in a given image are grouped into three regions by calculated local mean and variance values: stable region, neutral region, and complex region. To implement deinterlacing process, I use weight average filter by considering two factors (the likeness factor and the distance factor) and determine region characteristics. Simulation results inform that the presented method yields satisfactory results in objective metrics (PSNR, SSIM, implementation time) and subjective quality.