Particle Size Measurement Based on Image Multivariate Multiscale Entropy

An image-based multivariate multiscale entropy (MMSE) method used to measure the size of stacking particles is proposed. The intensive stacking particles with irregular shape and different sizes are collected online and processed by gradient image and one-dimensional coarse graining. Composite delay vectors are formed on row space to eliminate noise between multichannel variables. Then, the multivariate multiscale entropy is calculated in order to match the particle sizes. By comparing with the state-of-the-art methods, the experiment results indicate that the method can obtain the online intensive stacking particle size accurately without image segmentation.