Evaluation of color stability of ice cream with Spirulina platensis by image processing

The current study aimed to evaluate the color stability of ice cream samples with added Spirulina platensis powder by image processing. The change in color characteristics was monitored during 30 days of storage at -18°C. Images of the examined ice cream samples are processed with iceColor software to extract representative color for every sample under different lighting conditions. The used software implements three methods for color extraction: scalar-defined median color, vector-defined median color, and average color for pixels that fall in the selected area. All defined representative colors are compared versus instrumentally measured colors by Minolta chroma meter (CIE L*a*b* system) using correlation analysis. The results indicate that the highest correlation coefficient (about 0,99 as absolute value) is calculated for images captured using embedded flash. There is also a strong correlation between the other two lighting conditions (above 0,7). Based on the experimental results, it can be suggested that control of color changes during the storage period could be performed using image processing for images captured with embedded flash or without any additional lightning when the reduction of energy consumption is more important than precision of color measurement.