基于单义域邻接图的扫描图样矢量化,要生成具有较强整体性的条形域,包括线段、圆弧、箭头和多义线。它们的分类特征和分类知识均有模糊性,采用模糊推理进行分类。在分类时,以条形域较小和较大的直线度、圆形度和等宽度为模糊集合,以计算较少的升降半梯形为相应的隶属函数。这样分类符合人类认知方法,而且算法简单易于实现。 Based on the vectorization of the scan pattern of the single domain adjacency graph, a stripe domain with strong integrity, including line segments, arcs, arrows, and polylines, is generated. Their classification features and classification knowledge are ambiguous, using fuzzy reasoning classification. In the classification, the smaller and larger straightness, the circularity and the equal width of the bar domain are used as the fuzzy set, and the less lifting half trapezoid is calculated as the corresponding membership function. This classification is consistent with human cognitive methods, and the algorithm is simple and easy to implement.