In this article we discuss the possibility of using a conventional DSLR camera for color assessment of the prints enhanced with pearlescent pigments. Since these prints exhibit goniochromatic properties, color data were acquired in a multiangular manner and color estimation errors were assessed for the selected viewing angles. Colorimetric target-based camera characterization was performed by means of Artificial Neural Networks (ANN). In addition, ANN training was improved by implementing a multiobjective genetic algorithm with the aim to select the minimum number of different samples for the training set that will ensure efficient characterization. Our results indicate that the mean error of the performed characterization complies with the requirements placed on colorimeter in a print production. Furthermore, we show that the genetic algorithm optimization enabled an optimal training set selection for the given application, which makes the presented approach an efficient solution for multiangular color estimation.