Machine vision for size distribution determination of spherically shaped particles in dense granular beds, oriented to pelletising process automation.

Anibal Montenegro Rios, Damiano Sarocchi, Alejandro Lopez Valdivieso and Yuri Nahmad - Molinari.
Particulate Science and Technology, 29:356-367 (aceptado), 2011.


er Review Only A vision system for on-line determination of the size distribution of spherically shaped pellets was developed. Optimal ring illumination conditions for enhanced segmentation efficiency are inferred for deep granular beds. The performance of the system gives an accuracy in determining the size distribution better than 5% in static tests and for dynamic tests can be kept in the same range by setting the shutter speed and illumination intensities to proper values that depends on the exposure time due to blurring effects. The size distribution histogram given by the vision system is used as the input signal for controlling an electromechanical scanning device showing a high reliability for automation tasks. The system software was developed for open access on a commercially widely used platform in industrial environments in order to contribute to spreading out the use an implementation of machine vision systems to improve and optimize industrial processes. The proposed system is reliable, flexible, user friendly, cost effective, and easy to implement.