A novel digital imaging method for measuring cashmere color and its application

Abstract

The quality of cashmere, such as color and length, determines its price and application. In the current cashmere inspection system, color and length are tested by visual assessment, which is a subjective, time- and labor-consuming process. Herein, the goal of this research is to develop a new method of testing cashmere color using image analysis, and to study the application of color in length measurement. During the color measurement, cashmere was prepared under two sample placement methods, and color features including RGB, XYZ and Lab obtained by the new method were compared with the standard. The calculation method of optical index used in length testing was determined based on theoretical and experimental analysis. Experiments show that fixed weight and pressure are suited for cashmere color measurement. In RGB space, the correlation coefficients ( R2) between the two devices were calculated and were 0.990, 0.995 and 0.996 for parameters R, G and B, respectively. Good agreement also exhibited in XYZ space, with R2 equal to 0.994, 0.996 and 0.999 for X, Y and Z, respectively. This confirmed the accuracy of the proposed color measurement method in RGB and XYZ space. Finally, an accurate fibrogram was obtained by the proposed conversion model for calculating optical index from color values, which is the key curve to testing cashmere length. This study emphasis on methodological aspects and the results acquired are regarded as preliminary, as the experiments studied compose the first stage of research on the exploration of the application of image analysis on cashmere color measurement.