The Edge Processing On The Ray Testing Image Of The Pipeline

With the development of the country's project-gas transmission from the west to the east, in order to prevent the leakage of the gas pipelines, it is particularly important to test the quality of the welding seams. Since the pipelines pressures increase and the testing environment keeps changing, the traditional X-ray inspection method fails to fill the needs for testing. So researching a new nondestructive examination algorithm, and enhancing the testing and monitoring on the pipelines which are under construction and in service has a great significance for improving the security of the pipeline transportation.According to the feature of digital ray testing images of the pipelines welding seams, this paper finds that the images which are obtained by the image information input system always contain various kinds of noises and distortion. This phenomenon has severely affected the image quality. After analyzing the factors which affect the digital ray testing image quality, according to the feature of the ray inspection testing images, the paper selects the correlation parameters for the filter through doing experiments and sets up processing flow which is fit to the images. And then, it proposes a new edge detection algorithm which is based on the combination of frequency domain enhancement processing with Canny operator. Through deeply researching on the new edge detection algorithms, the paper realizes using the edge detection algorithms to complete the digital ray testing image processing which have typical defects. Through analyzing and optimizing the results of the processing, it finally achieves separating and quantitatively analyzing detects. Compare to the traditional edge detection algorithm, this algorithm is more effective to restrain noises, has superior performance on the edge detection accuracy, and has better edge extracting effect and edge continuity to complete the humidifying treatment of the images.At last, this paper checks the edge of the enhanced digital ray testing images, and find the edges which has typical detects effectively, then it proves that the level of sensitivity and definition of the enhanced images has obvious improvement. Finally. Mark out the edges which have the detective automatic recognition and assessment.