模糊数学在图像识别中的应用

Application of Fuzzy Mathematics in Recognition images

共晶碳化物分布的不均匀性是反映工具钢质量的重要指标之一。如何定量描述碳化物的不均性，如何客观地评定其级别，这是一个亟待解决的实际问题。本文应用模糊集理论，构造了三种评定碳化物不均匀性的数学模型，并编制了自动识别的计算机软件。通过实例对它们的优劣进行了比较，结果表明：模型 I 准确最高，为83%，模型 II 次之，为65%，而模型 III 仅为43%。

The unevenness of eutectic carbide distribution is one of the important factors that effect the quality of tool steel. How to describe the unevenness of carbide quantitatively and grade it are objectively actual problems that we have to solve.In this paper, based on fuzzy set theory, three mathematical models that can grade the unevenness of carbide, have been established. And we also develop computer softwares, which can identify them automatically. We contrast these three models by analyzing examples. The results indicate: model I has the highest precision reaching 83%, model II comes next amounting to 65%, while the precision of model III is only 43%.