Study on Digital Style Design

It has become difficult to differentiate industrial products as a point of view of the quality due to development of high production technology in each country and company. Development of products satisfying Kansei, requirement and preference of customers, has been becoming important and indispensable. This paper describes the development of a system emerging polka dots, which are widely used in the field of clothing pattern, based on Kansei of customers. The system is composed of two subsystems which determine color and arrangement of polka dots based on customer's Kansei. Questionnaires to investigate the relationship between the polka-dots colors, polka-dots arrangements and the impressions were performed. A neural network system to generate polka-dots design from required impressions was constructed based on the questionnaire results. The required impression is expressed by nine Kansei words with weight coefficients for color and by seven Kansei words with weight coefficients for arrangement.