Automated recognition of drill core textures: A geometallurgical tool for mineral processing prediction

To integrate the ore mesotexture into predictive block models, a core logging tool for automated textural pattern recognition is being developed. This paper presents the first step in this development: a methodology for the automated recognition of drill core textures. The proposed methodology is based on 2-D digital image analysis of drill cores. Texture information is extracted from digital images using gray level co-occurrence matrix (GLCM) and gray level run length matrix (GLRLM). Based on the information provided by these two methods, images were classified into six texture categories using multivariate discriminant analysis. A high classification success was obtained: 88% of the drill core images were correctly classified into their textural pattern category.