Ray Image Segmentation And Feature Extraction Technology Based On The Fuzzy Clustering

Ray detection is an important nondestructive detection method. In industry, materialinternal may be flawed, when defects in excess of the prescribed standards, will affect thequality of the products, seriously accidents may happen. X-ray detection can observe theinternal structure of materials, to determine the product qualified, plays an important role inindustrial production.Based on the X-ray real-time imaging equipment, this paperhas done research fortheX-ray imaging. From this we analyzed deeply in the characteristic of X-ray imaging, it iscomposedof the pretreatment, defectsegmentation and feature extraction.Because the effect of noise factors, the definition of the defect is not obvious, beforethe defect segmentation, should the processing for the X-ray imaging. This article adopts themethodof mathematical morphology to ray image preprocessing, the expansion and corrosionoperationcan meet the requirements of clarity to improve the defectdetails.On the basis of improving the clarity of the defect, the article carefully analyzed, weraise the method of the fuzzy c-means algorithm in the fuzzy clustering algorithm for imagesegmentation, under the condition of selecting reasonable parameters, effective divide thedefect area to the same area, highlight the defect targets, it can defect segmentation ofmaximumlimit, makes afoundation forsubsequentrecognition.The purpose of defect segmentation is defect classification, this paper studied thefeature extraction and analysis. Several types of features about location, contour, gray levelfeatures have been proposed and analyzed, which, according to some standard of defectdetection, lead to building the foundationof automateddetectionof the defects.