An application of food image enhancement method by using region selection in RGB color space

Proc. of 2017 International Workshop on Smart info-Media Systems in Asia (SISA2017), pp.83-86

SISA2017

Fukuoka

2017/09/07

C. Ueda, T. Azetsu, N. Suetake, E. Uchino

Food images looks delicious when warm hues are vivid because colors of foods mainly consist of them. However, the food images acquired in low illumination environments such as indoor places do not often give delicious impressions because of decrease in the intensity and saturation of colors. In this study, we propose an intensity and saturation enhancement method for food images by using a region selection in the RGB color space. In the proposed method, the intensity is first enhanced. Next, regions other than foods are marked by a user based on the intensity and hue conditions, and the saturation is enhanced only in food regions. The intensity and saturation enhancements are performed by the conventional method proposed by the authors.
To illustrate the effectiveness of the proposed method, experiments are conducted by using the proposed and conventional methods.