Image processing technologies nowadays are widely used in quality control. A new automated quality control system saves resources. Quality control in the fertilizer industry is carried out without interruption. Only indirect measurements can achieve the required quality production efficiency. This article presents the methodology for calibrating indirect particle measurement methods. Particle analysis is performed by processing digital images. Correction of results is carried out with a sieve correction factor. In this work, studies have been carried out using linear regression polynomial function and artificial neural network. With the latter, the best results are achieved with sufficient data for training.