Computer judgment of main color in a drawing for art psychotherapy assessment

Color in a drawing is one of the most important factors in art therapy assessments. The number and list of colors used, the area of each color painted, the blend of colors, etc., are common elements rated in the evaluation of art. The main color in a drawing is also a major concern in evaluation. This paper delineates the development of a computer system that judges the main color in a drawing by applying the methods available in the field of digital image processing. Often, the main color is a different one from the color with the largest area painted, which is usually a background color. The system devises an algorithm, a kind of artificial intelligence function, to judge the main color as human experts do based on the characteristics, number of pixels, and number of clusters in a drawing. The proposed system is verified through case studies, by comparing the judgments by the computer with those by the human raters. It can contribute to the quantification and objectification of human decisions, as well as reducing the time and effort of human raters, and can therefore be a useful tool and an aid for human experts in making decisions in art therapy.