基于x射线图像汽车轮毂缺陷自动检测与识别技术-automatic detection and identification technology of automobile hub defects based on x - ray images.docxVIP

优秀毕业论文精品参考文献资料Research on automatic detection and recognition technology of automobile hub defect based on X-ray imagesAbstractThe hub, an important force component of the vehicle, is directly related to traffic safety, especially in the case of the rapid rise in the number of motor vehicles, and the detection of the hubs is a big concern before they are out of the factory. Because the existing automatic detection software of wheel hubs is complex and poorly practical, based on the analysis of the wheel hub characteristic, such as image defects, this paper tries to develop the hub automatic defect detection and recognition systems which have been verified by experiment by means of the dynamic threshold image segmentation, defect feature extraction and BP neural network to achieve automatic detection and recognition system. Experimental results show that the system can detect and identify various defects of the hub casting, and particularly, it is able to handle the images having heterogeneity of product structure and complex background, what’s more the effect of detection and recognition is efficient, practical, and applicable to the needs of industries.In connection with the problem of large random noise and low contrast of the X-ray image, the integral noise reduction algorithm is employed to improve effective suppression of noise, and the contrast. Hub, however, is difficult to meet hub segmentation, (due to the limitations of the conventional algorithm of image segmentation) which raises the interesting region based on the hub image segmentation algorithm so as to solve the tough problem of the complexity of image segmentation. The experimental results have proved its effectiveness. Due to slow selection of feature transformation, it has been chosen that the changes in subprime search method are used to select the hub defect which effectively compress feature space, and lays the foundation for the type recognition of defects. Due to the common lower rate of BP n