ABSTRACT

Methods are described for the quantitative analysis of both the composition of sandstones and the two-dimensional morphology of their component mineral phases, based on image analysis of back-scattered electron (BSE) images of flat sections. Mineral phases are distinguished largely by their gray levels (intensities) in the BSE image, where contrast derives almost entirely from the compositions of the phases, although morphological criteria are applied where phases overlap in gray level. The concentration of each phase in each image is then measured as proportional area, and the average over many images gives an analysis of a significantly large area of the specimen. Morphological quantification is achieved, in a totally objective way, by operating a series of two-dimensional opening t ansformations on segmented images of the component minerals and pores, and the results are presented in the form of feature-size histograms, also summed over many image fields. Both the compositional and morphological analyses are capable of being fully automated.