Geological, textural and mineralogical classification of iron ores is routinely conducted by mine and exploration geologists on reverse circulation percussion, blasthole cone and bulk samples for grade control, defining ore and waste, determining lump:fines ratio and predicting downstream processing characteristics. Classification is based on physical hardness, colour, streak and other visual characteristics. However, logging is generally restricted to the five to 20 per cent of chips above about 2 mm. This means that most of the sample is too fine to be reliably logged for anything but colour or streak and the textural types must be inferred from other information. A recently developed software package called Recognition' allows automated identification and classification of iron ore textural types and gangue on a particle basis in polished section down to 0.01 mm using processed digital images. It is not designed to be used as a routine logging tool, but is applicable where higher definition is required at the resource evaluation stage on selected percussion, drill core or bulk samples. Recognition' processes data obtained after optical image analysis of fine iron ore samples. It contains several parts, like Recognition', Composition', Liberation by Total Iron', Liberation by Phases' and Statistics', which allow vast and diverse information to be obtained about the iron ore samples being studied._x000D_
Identification and classification of the textural type of each particle is performed according to the CSIRO-Hamersley Iron Ore Group Classification Scheme. The recognition procedure encompasses a good knowledge of different ore types and a logical knowledge tree, managing fuzzy boundary problems and the potential for a few different ore types in the same particle to be considered. This new technology introduces strict criteria to make the process of identifying particles more objective. Automated recognition standardises the identification procedure and significantly saves time. It enables users without a strong background in mineralogy to perform accurate texture identification._x000D_
The Recognition' package has a modern user-friendly interface and Word compatible output. The combination of all the features mentioned makes the package a convenient tool for detailed studies of fine iron ores.