A Rule-base System for Image Classification

This is the project that I did around 1987-1988 at Chinese Academy of Forestry for my Master's degree in remote sensing. A paper describing this work was published in Forest Research, a professional journal in Chinese. Soon after coming to U.S. in 1990, I put together an article in English as a rewritten version of the paper with the title of "A study on expert system aided forest and landuse classification of remotely sensed images in north China".

ABSTRACT In order to incorporate various ancillary data into the procedure of remote sense image classification, the technology of expert system is investigated as a tool of information integration, using evidential calculus proposed by Shafe. The design and implementation of an expert system for forest and landuse image classification is discussed. A new algorithm, called Fuzzy Supervised Classification (FSC), that works in conjunction with the system is also developed. The whole system is applied to an image in an experimental area in North China and the initial results are reported.

The English translation of the paper in full length can be downloaded in PDF format, which can be read with Adobe Acrobat Reader.

The following figure shows the general structure of the system. Module-A and Module-C were written in Fortune, while Module-B in Lisp.

Home Page of W. Zhou Scientific/Engineering Projects Gallery of Scenic Photography