A Colour Perception Detection Model Based on The Spectral Characterictics

Abstract

Information needs for one's color perception are needed in the fields of medicine, engineering, astronomy, biomedicine and so on. The demand for accurate assessment of color perception must be met by the perception detection tool used. Ishihara's test, as a perception detection tool that is still used today has insufficient accuracy. This research aims to create a system that can detect a shift in one's color perception, relative to the average color perception of a number of respondents. Through plotting the respondents' perception points, in the CIE coordinate system (Commission International de I'Eclairage) XYZ can be calculated the average euclidean distance, ED, relative to the reference point and the distribution of x and y groups of perception points around the point of reference. Both size, euclidean distance and distribution are used as indicators of average color perception so that an assessment of one's color perception is given based on the results of comparison between color perception points and color perception indicators. The tool used to do the test is Delphi version 7.0 software. the research material used is the RGB (Red, Green, Blue) color image format. The results of a person's color perception study are divided into three levels, namely: (1) "normal" assessment if euclidean (ED) perceptions are smaller than the euclidean (ED) average (2) the "somewhat normal" assessment if the distribution of x and y is smaller rather than the color of perception and the distribution of x and y (3) the assessment is "abnormal" if the color of perception is greater than the max distribution of x and y. A new perception point assessment that is in level one is used to up-date prevailing perception indicators. Up-dating condition constraints affect the quality of the threshold average perception specifically and the quality of the results of the perception detection system in general.