An Affective Analysis of the Compositions of Digital Photo

The advances of image processing technology have enabled digital photographers to digitally correct common photography problems, e.g. under exposure, colour deviations, or low contrast. Most off-the-shelf image processing software packages also integrate automatic functions to assist the casual digital photographers on the correction tasks. Image quality enhancement has also been one of central research themes in image processing area. In recent years, there have been quite a few studies focusing on content-based image processing techniques. Many methods have been proposed to support image content editing tasks and content-sensitive operations. However, automatic techniques on processing image content are still a problem for the researchers to tackle. This project carries out an affective study on scene compositions. Composition-relevant guidelines and affective adjectives will be collected from design literatures. Semantic differential rating experiment will then be carried out to investigate the mappings between the scene compositions and the viewers’ affective responses. This interdisciplinary study combines design knowledge and computational techniques to support casual digital photographers on composition editing tasks. The outcome of the project can potentially be published in the related international journals and conferences. On the practical side, the can be fully developed as software functions to facilitate automatic composition enhancements or media design education tools.

Project ID:PE10010-0049
External Project ID:NSC100-2410-H182-026