• DocumentCode
    1982882
  • Title

    Identification of factors related to the enhancement of image-quality for subjective image-quality assessment model based on psychological measurement

  • Author

    Aiba, Eriko ; Numata, Koyu ; Fujisawa, Takashi X. ; Nagata, Noriko

  • Author_Institution
    Grad. Sch., Kwansei Gakuin Univ., Sanda, Japan
  • fYear
    2011
  • fDate
    19-21 May 2011
  • Firstpage
    174
  • Lastpage
    177
  • Abstract
    The purpose of this study is to identify the physical factors affecting image-quality for a comprehensive enhancement in subjective image-quality assessment by human being. In recent years, digital cameras have become extremely popular because they enable image recording with very high fidelity and are affordably priced. However, in subjective image-quality assessment by human, one does not always consider that images with high fidelity are of high quality. In this study, a psychological experiment was performed as the first step to develop a subjective image-quality assessment model that considers kansei of human being by using conjoint method. We particularly focused on the identification of physical factors without depending on a scene. The results of this experiment showed that “luminance” and “contrast” may contribute to an enhancement in the image-quality assessment, regardless of the type of scene used. It was also suggested that human memory, in which the colors preferred by the individual are retained, also plays a major role in a subjective image-quality assessment.
  • Keywords
    image enhancement; conjoint method; image-quality enhancement; physical factor identification; psychological measurement; subjective image-quality assessment model; Bridges; Digital images; Humans; Image color analysis; Object recognition; Psychology; Skin; Image-quality; Kansei; Physical factor; Psychological factor; Subjective assessment;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Human System Interactions (HSI), 2011 4th International Conference on
  • Conference_Location
    Yokohama
  • ISSN
    2158-2246
  • Print_ISBN
    978-1-4244-9638-9
  • Type

    conf

  • DOI
    10.1109/HSI.2011.5937362
  • Filename
    5937362