Determination of L*a*b* Parameters from RGB of Digital Image by Application of IMG-Pardazesh Color Measurement Instrument

Document Type : Research Article

Author

• Seyed Ali Jafarpour

Sari University of Agricultural Sciences and Natural Resrouces

https://doi.org/10.22067/ifstrj.v12i5.43517

Abstract

Introduction: The use of computer vision technology has been highly successful in food classification in the past and it has continued this success in recent times. However, a number of opportunities to progress computer vision technology exist which are critically examined based on cost and feasibility. A range of hardware options are considered along with a range of software options. The economic cost of implementing new hardware continues to prove a major impediment. Thus future efforts need to be focused on maximizing the potential benefits of the existing hardware framework and instead concentrate on developing improved software. Of the improved software available the aspect that offers the greatest promise is more efficient analysis of food surface texture attributes which will lead to more powerful understanding of the relationships between quality factors and experimentally measured food quality.

Materials and Methods: In this study, the efficiency of IMG-Pardazesh instrument in color measurement in comparison with CIE L*ab, Hunterlab and Patch Tool color systems was evaluated. The IMG-Pardazesh instrument was designed and manufactured based on CIE 45/0 standard and all measurements were performed based on the ColorChecker® 24 Patch Classic target which is an array of 24 scientifically prepared natural, chromatic, primary and gray scale colored squares in a wide range of colors. Many of the squares represent natural objects, such as human skin, foliage and blue sky. Since they exemplify the color of their counterparts and reflect light the same way in all parts of the visible spectrum, the squares will match the colors of representative sample natural objects under any illumination, and with any color reproduction process.

Results and Discussion: According to the results, the regression value (R2) of L*a*b* resulted from IMG-Pardazesh compared to CIE L*ab recorded at 0.996, 0.998 and 0.980, respectively. In comparison withHunterlab, the values were equal to 0.983, 0.981, 0.871, and compared to Patch Tool system were 0.935, 0.881 and 0.953, respectively. However, to base an unbiased conclusion it is necessary to consider the numeric value of data that can be calculated in form of Root Mean Square Deviation (RMSD) rather than the similarity of color changes pattern.Therefore, as much as the RMSD value becomes smaller, the validity of color measuring instrument become greater compared to the standard system. RMSD was calculated following below formula:
RMSDL= √((∑_(i=1)^n▒(L_i^*-L_(p )^* )^2 )/n)
RMSDL= √((∑_(i=1)^n▒(a_i^*-a_(p )^* )^2 )/n)
RMSDL= √((∑_(i=1)^n▒〖(b_i^*-b_(p )^*)〗^2 )/n)
Which Li*,ai*,bi* are color parameter from Patch Tool system and Lp*,ap*,bp* are color parameters from other color systems.
By calculating the RMSD index, it was revealed that numeric value of L*a*b* from IMG-Pardazesh was slightly lower than that of CIE L*ab. Compared to Hunterlab system, apart from a* value, the RMSD was remarkably lower in L* and b* values. By calculating the normalized error of means (e), the values of eL, eaandeb from IMP-Pardazeshwereequal to 0.776, 1.184 and 0.968, respectively, whereas, the same parameters for CIE L*ab were recorded as 0.882, 1.243 and 1.124, respectively, and for Hunterlab system were found to be 1.085, 0.933 and 1.423. Furthermore, computingthe average normalized error of means (e ̅) in CIE L*ab compared to L*a*b* from IMG-Pardazesh indicated that all color parameters had higher total average error and it terms of Hunterlab again L* and b* showed higher error. In a study conducted by Mendoza et al. (2006) on application of image analyzing for evaluation of food items color, the authors stated that that sRGB standard (linear signals) was efficient to define the mapping between R′G′B′ (no-linear signals) from the CCD camera and a device-independent system such as CIE XYZ. The CVS showed to be robust to changes in sample orientation, resolution, and zoom. However, the measured average color was shown to be significantly affected by the properties of the background and by the surface curvature and gloss. Thus all average color results should be interpreted with caution. L*a*b* system is suggested as the best color space for quantification in foods with curved surfaces. In another study on evaluation of L*a*b* units from RGB parameters, Leon et al. (2006) presented five conversion models as: linear, quadratic, gamma, direct, and neural network. Additionally, a method was suggested for estimating the parameters of the models based on a minimization of the mean absolute error between the color measurements obtained by the models and/or usinga commercial colorimeter for uniform and homogenous surfaces. In the evaluation of the performance of the models, the neural network model stands out with an error of only 0.93%. the same authors also stated that on the basis of the construction of these models, it is possible to find a L∗a∗b∗ color measuring system that is appropriate for an accurate, exacting and detailed characterization of a food item, thus improving quality control and providing a highly useful tool for the food industry based on a color digital camera.In conclusion, the IMG-Pardazesh instrument have lower error in determination of L*a*b* parameter from RGB of digital image compared to the other tested systems.

Keywords

• Digital Image
• Color Measurment
• RGB
• L*a*b*

References

Francis, F. J., & Clydesdale, F. M. (1975).Food Colorimetry Theory and Applications.,: Westport.

Jackman, P., & Sun, D.-W. (2013). Recent advances in image processing using image texture features for food quality assessment. Trends in Food Science & Technology, 29(1), 35-43.

Leona, K., Meryb, D., Pedreschic, F., &Leonc, J. (2006). Color measurement in L*a*b* units from RGB digital images. Food Research International, 39(10), 1084-1091.

Mendozaa, F., Dejmekb, P., &Aguileraa, J. M. (2006). Calibrated color measurements of agricultural foods using image analysis. Postharvest Biology and Technology, 41(3), 285-29

Nobbs, J.H. & C. Connolly (2000).Camera-based colour inspection Sensor Review. MCB University

Press, 20.14-19.

Tanska, M., D. Rotkiewicz, W. Kozirok, & I. Konopka (2005).Measurement of the geometricalfeatures and surface color of rapeseeds using digital image analysis. Food Research International, 38,741-750.