The sensory evaluation of food products by consumer panel is commonly performed on physical samples. However, the evaluation of visual attributes like color can be simplified using digital images. In this work, color acceptability (CA) of skimmed goat's milk yogurt (GMY) enriched with lucuma (Pouteria lucuma) pulp (LP) and yacon (Smallanthus sonchifolius) extract (YE) was optimized using RGB images. For this, Simplex-Centroid Design (SCD) of Response Surface Methodology was used to explore 10 formulations that consisted of mixtures of GMY (82-89%), LP (10-17%) and YE (1-8%). A computer vision system was used to acquire RGB digital images of the samples. The images were used in an online survey to 100 consumers for the evaluation of CA with a structured 9-point hedonic scale. Additionally, the instrumental color (RGB components) of samples was measured from the images using ImageJ software. Statistical analysis of the SCD showed that cubic models satisfactorily explain the effect of GMY, LP and YE on CA and RGB components (p<0.05 and R2>85%). The analysis of the response surface of CA model showed that the percentage ranges that optimize the CA of the yogurt were: 83.96-85.85% of GMY, 10-10.21% of LP and 4.15-6.04% of YE. Additionally, Desirability Function revealed that the specific values that optimized the CA were: 85.01% of GMY, 10% of LP and 4.99% of YE. The models obtained for RGB components allowed to calculate that the yogurt with optimal CA has a color composed by R = 221.24, G = 196.43 and B = 147.60. Thereby, the RGB images were useful in the optimization of the CA of the yogurt and in the study of the effect of formulations in the components RGB. This procedure that utilize a practical and inexpensive instrumental support could be used in others food products that have uniform color.