AbstractThis paper introduces a Prolog‐based support system for use in developing a picture recognition program. Earlier, an image recognition program was introduced using such items as cigarette packages, tableware, etc. However, because of the disadvantages of that program, a new one was necessitated.Prolog and Lisp, languages designed for use in artificial intelligence, are more suitable for use in high‐level processing, e.g., image recognition and understanding. However, Prolog shows greater potential for use in writing routines in the recognition process. But Prolog is not very effective in handling numerical calculations involving large amounts of two‐dimensional images. As a result, Prolog systems are examined herein that have been developed using additional features that allow for implementation of other languages and databases.Here, the PROFIER (PROlog system with Fortran as Image analyzER) system is described. In this system, the knowledge and control structures are written in Prolog, and for raw data processing and image feature extraction, processes that are difficult to handle with Prolog, the system is expanded for programming in FORTRAN. The configuration and characteristics of PROFIER are described, and the performance is eval