Principal component analysis for characterising homogeneity in powder mixing using image processing techniques

The many methods which exist to characterise the quality of a powder mixture have been recently reviewed and linked with mixing mechanisms in current literature. In this paper, we try to develop a novel methodology for defining and characterising homogeneity using principal component analysis (PCA) as an alternative to well-known statistical methods, such as auto-correlation functions or variances. We apply this to image analysis for the case of a powder mixture flowing out of a continuous mixing device. An emphasis is placed on the calculation in real-time of the degree of homogeneity of loose materials on the conveyor belt, carrying this mixture. Mass flow disturbances applied to a binary mixture are studied by the proposed methodology, which is found to be sensitive to small structural defaults.