Incoherent material monitoring by image analysis for mineral processing

The study of a deposit of incoherent materials can be carried out by chemical and mineralogical analysis, or, adopting a new approach, by image analysis of surface properties. The aim of this methodology is to establish a link between image parameters (texture and colour), defined on the basis of optical properties, and physico-chemical parameters characterising the set of particles to be evaluated. The technique was applied to an ore deposit of mineral sand composed by three different lithotypes. An appropriate number of samples for each lithotype was selected. Then, a vector of attributes (pattern vector), composed by either texture or colour parameters was associated to each sample. Image parameters are quite sensitive to the conditions in which the images were acquired: optical properties may be strongly influenced by physical condition, in terms of water content and illumination conditions, and by image acquisition accuracy, such as a correct focal point. The standard conditions in acquiring image were chosen in relation to the conditions found on the deposit during sampling. To verify the reliability of the proposed methodology, 45 images were acquired under different conditions of humidity, focusing and illumination. The textural analysis was arranged on these images to evaluate the influence of this parameters on image properties. The data acquired by image analysis are to be used for remote controlling of feeding in mineral processing plants.