The plots in Fig. 5 and Fig. 6 indicate that the colorimeter is very dependent on the scattering properties of the sample. The processed meat products have to some extent been homogenized due to the preprocessing, and they are visually less glossy than the fresh meat samples. This glossiness can lead to specular reflectance which will lead to a less significant contribution to the chromatic components a* and b* ( Kigle-Böckler & Hammond, 2012). The results indicate that the diffuse illumination by the VideometerLab gives less dependency on the lightness and glossiness of the sample than the simple illumination and filtering employed by the colorimeter. These results are in accordance with the results of the study on salmon meat by Yagiz et al. (2009). This study on different meat types and processing levels reveals that the reflectance properties of fresh meat, whether that being poultry or livestock animals, can influence the colorimeter measurements and that a diffuse illumination and multispectral imaging of the sample can be a way of overcoming this problem.
Imaging the samples has several advantages compared to the colorimeter measurements. First of all it captures the color variation across the sample, but as stated above, the illumination of the imaging system is preferable to that of the colorimeter. The multispectral images also enable for several other analyses than color, such as finding fat and lean meat ratio measures by segmentation or spoilage detection. Examples of such uses are explained in Trinderup, Dahl, Carstensen, Jensen, and Conradsen (2013) and Dissing et al. (2012).