Photo quality enhancement by relocating subjects

• Published in: Cluster computing Vol. 19; no. 2; pp. 939 - 948
• Main Authors: Jeong, Kyuman, Cho, Han-Jin
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.06.2016; Springer Nature B.V
• Subjects: Algorithms; Composition; Computer Communication Networks; Computer Science; Digital cameras; Digital imaging; Operating Systems; Partial differential equations; Processor Architectures; Relocation; Photography; Subject relocation; Quality enhancement
• ISSN: 1386-7857; 1573-7543
• DOI: 10.1007/s10586-016-0547-z

Photo quality enhancement by photographic evaluation is a challenging problem in photography. In this paper, a new photo quality enhancement algorithm for photographic recomposition based on subject relocation is presented. Most related research focuses on image size alteration, or cropping, while the proposed algorithm in this paper is extending existing ones by relocating foreground object to meet one of the composition rules. The contribution of this paper is to overcome the problem of existing algorithms which cannot place the main object according to photographers’ intentions. In detail, the presented algorithm is consist of three parts: image matting, region filling and image recomposition. Image matting is used to extract foreground object from background image. The region where foreground object is separated must be filled naturally without any seams or artifacts by the aid of region filling algorithm. Finally, the foreground object can be relocated to a proper location with a better composition score. Experimental results will prove the excellence of the presented algorithm. The presented method can be embedded on commercial cameras, so the authors expect that it can be applied in the near future.