Research advances in geospatial automated image analysis tools and feature extraction algorithms have matured in recent times to levels of practical applicability. The consolidation of such tools and algorithms would result in enhanced image analysis capabilities. This has motivated research in developing formalisms for representation of process information that can assist in integrating tools and data sets, and allow for effective dissemination of research results across a raster-rich geospatial distributed information system. We propose a framework for a formal representation of image processing operations and algorithms as process metadata with generalized adaptability across all raster data sets. The elements of such a formal representation and their significance in the domain of distributed information systems are presented.