RGB camera-based image technique for color measurement of flavored milk

Abstract

The potential of RGB cameras to extract and measure the color of flavored milk was assessed using the image analysis technique. A simple procedure has been proposed using RGB camera for the color measurement of flavored milk samples. The color measurement algorithm involved the transformation of device-dependent RGB color values into device-independent CIE L*a*b* color values. The study helped to determine the variation in colorimetric results when a computer vision system (CVS) was evaluated using five different cameras (C1-C5). CCD (charge-coupled device) and CMOS (complementary metal-oxide-semiconductor) cameras yielded similar results. The variability in colorimetric results due to the effect of light source and camera specifications can be minimized by a combination of calibration, color measurement, and correction algorithms. The total color difference (ΔE) was in the range of 0.20 – 6.42 and can be used as a parameter to characterize cameras for color measurement accuracy. C1-C4 based CVS accurately measured color compared to C5. The study established that the computer vision method based on RGB camera is a useful tool for rapid color assessment of flavored milk samples.