Morphology characterization of irregular particles using image analysis. Application to solid inorganic fertilizers
2018
Abstract This research deals with how to characterize the morphology of mineral fertilizers using dynamic image analysis. A machine-vision system was first developed to capture digital images of irregular particles. The vision system was designed as a mechanical assembly to generate a flow of separated grains passing in front of a camera, combined with a flash. The vision system’s parameters were then calibrated and optimized – particularly in terms of image resolution, light sources and exposure time. An optimal value of 34 pixels/mm was obtained for the image resolution. Then, the system was tested and validated by imaging perfect plastic spheres of known size. Secondly, an image-processing algorithm was developed to extract geometrical and morphological information for various particles. Various shape parameters, describing the differences of a particle from a reference geometric form (e.g. a circle or ellipse), were calculated as outputs of the image-processing treatment. Statistical analysis was then applied to determine the convergence of shape parameter distributions and also the repeatability of measurements. 45 fertilizers, with grain shapes ranging from highly irregular to nearly spherical, were prepared and imaged. For all these fertilizers, the distribution of shape parameters was quantitative, representative, repeatable and reproducible using the machine-vision system developed. In order to investigate the best parameters for characterizing particle morphology, statistical correlation was applied to deduce a list of independent shape parameters. These parameters were relevant and could further be used to characterize the morphology of any fertilizer. The independent parameters thus obtained were subsequently used to detect the correlation between morphology and distance traveled by fertilizer particles thrown out by a centrifugal spreader. Experimental tests were conducted using the CEMIB device to determine fertilizer spread pattern from spinning discs. Results showed that, within the same range of size, mass density and spreader operating parameters, spherical and rounded particles traveled further than elongated and angular particles. The angularity index parameter, denoted by ANGInd (which characterizes whether a particle is rounded or angular), showed most potential application to explain the aerodynamic behavior of irregular particles spread by spinning discs.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
31
References
4
Citations
NaN
KQI