Study On Rapid Identification Of Defects In X-Ray Digital Imaging Detection

X-ray digital real-time imaging technology as a new technology in the field of welding defect detection, have been introduced into the pressure vessel, pipeline and equipment quality inspection and supervision. However, in recent years, domestic equipment upgrades and new putting-in-service proactively, there is still a image processing system was not equipped with or not perfect, in the process of testing personnel dependence is big, can’t timely and accurately identify, to elaborate target data processing is slow, the function is too single, human-computer interaction problem such as unfriendly. To solve these problems, this paper carried out a number of key technology research.X-ray digital imaging is studied at home and abroad, the application and development of the various periods, the contrast analysis of the conventional X-ray test and X-ray digital method, principle, system structure, the characteristics of inspection process and its application range and so on, puts forward two methods of their advantages and limitations, and believes the film method still has great advantages and application of the space, but the intelligent, highly efficient of the real-time imaging detection and standardization is the inevitable trend of development.Comparative study in this paper the geometrical transformation, gray level transformation, histogram equalization processing, based on image airspace gradient sharpening and Laplacian sharpening enhancing processing method. For crack defects images is proposed based on two kinds of the airspace image processing and segmentation method, suggests that these two schemes can be better to remove noise and maintain the edge of the weld defects and details. In the study of fast recognition algorithm, introduced the gray analysis method, using column grayscale change characteristics of signal singularity in the defect area, put forward a kind of based on feature analysis in column gray rapid identification method of X-ray digital welding seam images, confirmed the method combined with pretreatment method, the possibility that can be used to quickly identify. Recognition algorithm of defect areas were studied, put forward the regional defect quickly identify positioning and automatic segmentation method. Experiments show that the method can quickly locking defect area, narrowing the scope of evaluation effectively. In addition, the use of MATLAB software GUI ray digital image is developed fast identification of weld defects of human-computer interaction platform.