Research On Method Mapping Of Material Appearance Corrosive Low-level Features To High-level Semantics Based On Fuzzy Domain Ontology

It is impossibly for designing products not to relate to the selection of material.Testing the adaptability of the material is to check its survival ability in the environment. Inthe environmental testing, due to the effect of various environmental factors, the surface ofMaterial will has the corrosion characteristics, mainly about: cracking, bubbling, pitting,spalling, discolor and so on. According to the corrosion characteristics, it can confirmOutward Appearance Corrosion Rank of the material. The most direct way to describe thematerial’s outward corrosion information is from the status of material outward corrosion.How the material corrosion appearance is not only one of the most important parts, but anindispensable basis for product designing and materials selection.Handling of the material outward corrosion characteristics from the underlyinginformation to the interpretation of the high-level semantic description exists a "semanticgap" problem in the present situation of computer information. How to accurately identifyand express material outward corrosion characteristics from the high-level semantic is needto be solved in the application of the computer information. In order to reduce the"semantic gap" at present, on the one hand, to build a description model, which can fullydescribe the underlying material outward corrosion characteristics and high-level semanticfeature, to build a bridge between low-level features and high-level semantic, on the otherhand, to construct a classifier to implement the mapping of material outward corrosioncharacteristics to the high-level semantic feature.It is difficult to understand the image directly obtained from the low-level visualfeatures, which requires analyzing the semantics of the image deeply, making the computercan understand the similarity based on human cognition. To this end, we associate the body,proposed framework for ontology-based image: by ontology-based semantic extension,retrieval semantic query process to make up for the lack of information; through theontology, the semantic definition of the concept of image relationship. There is often aambiguity between the semantic concept description, so Fuzzy set added to the ontologytechnology to build the FDO (fuzzy domain ontology). Image fuzzy domain ontologymodels include the visual image features, high-level semantic concept, not only take fulladvantage of the low-level features of image itself, but also consistent with the visualimages of people from understanding, thus filling the "semantic gap" between the low-level visual features and high-level semantics.In order to achieve high-level image automatic extraction of semantic features, thispaper introduces the basic theory of SVM, support vector machine using the underlyingcharacteristics of the image mapped to the ontology of the high-level semantic concepts, sowe will automatically obtain semantic annotation information of images from theunderlying characteristics. The SVM classifier is a absolute classifier, but a sample can notbe completely attributed to a class, we constructed fuzzy support vector machines, theshortcomings of fuzzy support vector machine is it can’t correctly determine the sampleweight factor. so we introduce the vague sets and eventually constructed based vagueFSVMs (fuzzy support vector machine) classifier. In order to better improve the low-levelfeatures to high-level semantic mapping. Finally, under the given experimental samples,carried out in certain verification.