Thin sections carry considerable precise quantitative information concerning physical properties of the rock such as permeability, log response, and acoustic phenomena. This information is not obtained through conventional petrographic procedures such as point counting or simple linear measurements made with a graduated ocular. Instead, multispectral-image-analysis techniques are required. Critical parameters of such a system include: (1) selection of correct magnifications; (2) selection of proper spectral bands; (3) construction of a high precision digital filter able to define precisely grain and pore boundaries at the upper surface of the section; (4) identification of an algorithm to express in useful fashion the complex geometries of grain and pore as well as correc ly identify minerals and mineral phases; and (5) development of software to analyze such data. Output data is copious--for example, more than 80 variables are used to describe pore size and geometry. The result, however, is precise estimation of such properties as permeability and identification of pore characteristics associated with enhanced or reduced permeability. The procedure at present requires minicomputer-level hardware for optimal data acquisition and processing.

End_of_Article - Last_Page 473------------