The Application Of Image Processing Techniques In Welding Defect Detection

During the image acquisition by the welding defect detection system, the welding defect images will have some quality problems such as the random noise generated by the image acquisition equipment, the fuzzy character and edge of the defects caused by low-contrast. All of the problems are due to the influence of the external environment,and the existence of these quality problems will impact the normal identification of the welding defect. Therefore, to meet the need of correct judgement and identification for welding defects,a certain image processing technology must be adopted to remove the image noise, increase the image's contrast and improve the visual effects of the images.In this paper,by the actual situation of the project, a approach of preprocessing system for welding defect image is proposed and completed after searching related pre-processing information and relevant equipment. Through the research of image acquisition method,we choose the acquisition equipment of weld defect image, then complete the common welding defects in the work of the image acquisition.In this paper,according to the characteristics of welding defect images, the contents and methods of image preprocessing are chosen and Matlab tool is used for experiment contast and verification by using the methodolodgy of image grayscale, image enhancement, image smoothing, image sharpening, etc.,the algorithm and the relevant parameters of image pre-processing are designed and determined, the design work of the image weld defect preprocessing system is completed.At the same time,under the WindowsXP environment, the pre-processing system was developed to achieve the functional by using the Microsoft Visual Basic 6.0 Language. This system makes the Matlab M-Files translate and edit to the.EXE file directly to achieve the seamless integration comband VB and Matlab.Through the processing of the welding defect image pre-processing system, we get welding defect images with less noise disturbance, higher image contrast and better visual effects by the applicated images from this system,and provide the security for the identification of welding defects.