声学地层剖面记录的图像识别研究之二（墓于模糊集理论的专家系统方法）

Pattern recognition of acoustic sea-bed profiling records (part 2: a fuzzy logic reasoning expert system)

本文讨论应用基于模糊集理论的专家系统方法进行声学地层剖面记录的图象识别问题。该方法在提取声学地层剖面记录所表现的各个图象特征状态以后，逐个计算该特征状态集对不同地质属性的地层图象的标准特征状态集的贴近度，再按照贴近度最大值作出所探测地层的地质类别的判断。

An expert system based on the fuzzy set theory has been developed for the computer-basedpattern recognition system of Acoustic Sea-bed Profiling Records. The standard pattern characteristicstate sets appearing on the ASPRs corresponding to 10 geological categories of marille sediment layershave been defined by experienced specialists,and the weight of each pattern characteristic-state in determining the geological classification calculated through a detailed analysis. Through inquiring aboutthe pattern characteristic-state appearing on an ASPR being interpreted and then calculating the anproximating factors of this ASPR to the standard ones corresponding to the 10 categories of sedimentlayers respectively, the geological classification of the sedimellt layer can be determined by selecting themaximum approximating factor.