METHOD FOR CLASSIFYING SHAPE IMAGES BASED ON THE TOPOLOGICAL THEORY OF PERCEPTUAL ORGANIZATION

摘要

A method for classifying shape images based on the topological theory of perceptual organization includes steps of: extracting edge points of the shape images (S1); building a topological space, and calculating the expression of the extracted edge points in the topological space (S2); extracting global features according to the expression of the edge points in the topological space (S3); extracting local features according to the expression of the edge points in the Euclidean space (S4); merging the global features and the local features, and adjusting the weight of the local features during a merging process according to the match degree of the global features (S5); and classifying the shape images according to the merged features (S6). The present invention is applied to an intelligent visual monitor system to assist the monitor system to classify the objects in a scene, so as to make the monitor system really understand what is occurring in the scene and employ different security levels according to the different object classifications. It is applied to an automatic driving system for determining the classifications of traffic signs, so as to let the automatic driving system become more intelligent.