A study on picture feature extraction and picture recognition

In general, the feature extraction process in picture recognition requires troublesome global picture computations. The purpose of this paper is to describe a practical and general purpose feature extraction algorithm for binary pictures by utilizing perceptions. This algorithm proves that effective feature values can be extracted on the basis of fully local picture computation. The localness of a picture computation is evaluated by the coacept called the “order” of a perceptron. Further, based on this feature extraction algorithm, an efficient picture recognition system is realized. The effectiveness and practicability of this system are clarified by computer simulation experiments.