Research Of Detection Technology Based On X-ray Image Of Ammunition Defects

As an efficient and reliable non-destruction technology, industrial X-ray inspection is applied widely in different industrial fields, especially in the fields of national defense and aviation industry.The projectile, as a military hardware, is so dangerous during themanufacturing and applied process. So the figure of the projectile and the quality of the ammunition must accord with the general specification. Guaranteeing the quality and reliability of the projectile availably is the main purpose of this project. During the manufacturing process, detecting defects, such as the gap, is more important. So defects detection on ammunition based on the image of the X-ray projectile is the emphasis of this thesis.This thesis researches on the qualitative detection and the quantifiability detection of the projectile. Finding types and location is the purpose of qualitative detection. In this part, it does research on image pre-processing and image segmentation. In the image pre-processing, the image filter and enhancement technology are presented. Emphasis is put on the study of the image segmentation. In the edge detection, in the light of the feature of X-ray image and the ammunition which is inserted, combining with wavelet transform, and comparing with the traditional edge detection algorithm, this thesis presents a method of defect edge detection that adapts to the X-ray projectile image. Finding defects'extent is the purpose of the quantifiability detection, and this thesis does research on the gap quantifiability.