The quality control of premix components is a crucial stage of their production. It directly influences a number of final product characteristics, including homogeneity, granulometric properties, and the effectiveness of the premix ingredients. This study highlights the relevance of using computer vision based on the OpenCV library for image processing of salt samples, as well as the application of deep learning with TensorFlow for automated quality classification. The results indicate whether the proposed tools are suitable for an efficient quality control process, enabling a reduction of human factor influence and improving the accuracy of defect detection.
Kiktev et al. (Wed,) studied this question.