In industrial X-ray inspection, in order to identify weld defects automatically, raise the identification ratio, and avoid processing of complex background, it is an important step for sequent processing to extract we...In industrial X-ray inspection, in order to identify weld defects automatically, raise the identification ratio, and avoid processing of complex background, it is an important step for sequent processing to extract weld from the image. According to the characteristics of weld radiograph image, median filter is adopted to reduce the noise with high frequency, then relative gray-scale of image is chosen as fuzzy characteristic, and image gray-scale fuzzy matrix is constructed and suitable membership function is selected to describe edge characteristic. A fuzzy algorithm is adopted for enhancing radiograph image processing. Based on the intensity distribution characteristic in weld, methodology of weld extraction is then designed. This paper describes the methodology of all the weld extraction, including reducing noise, fuzzy enhancement and weld extraction process. To prove its effectiveness, this methodology was tested with 64 weld negative images available for this study. The experimental results show that this methodology is very effective for extracting linear weld.展开更多
In this paper,a photoelectric device is introduced,which is used in detecting the quality on internal surface of thin and long steel pipe.In this device,the CCTV lens is used for extract tile flaw information on inter...In this paper,a photoelectric device is introduced,which is used in detecting the quality on internal surface of thin and long steel pipe.In this device,the CCTV lens is used for extract tile flaw information on internal surface of the pipe,and make IBM-PC/AT 486 computer as controlling and image processing system.By this instrument,the functions,such as the digital conversion of input information,image processing,classification of recognition and output display can be obtained.In the petroleum and chemical industry,by using this apparatus,we can detect the quality on internal surface of various metal pipes with real-time automatically.展开更多
文摘In industrial X-ray inspection, in order to identify weld defects automatically, raise the identification ratio, and avoid processing of complex background, it is an important step for sequent processing to extract weld from the image. According to the characteristics of weld radiograph image, median filter is adopted to reduce the noise with high frequency, then relative gray-scale of image is chosen as fuzzy characteristic, and image gray-scale fuzzy matrix is constructed and suitable membership function is selected to describe edge characteristic. A fuzzy algorithm is adopted for enhancing radiograph image processing. Based on the intensity distribution characteristic in weld, methodology of weld extraction is then designed. This paper describes the methodology of all the weld extraction, including reducing noise, fuzzy enhancement and weld extraction process. To prove its effectiveness, this methodology was tested with 64 weld negative images available for this study. The experimental results show that this methodology is very effective for extracting linear weld.
文摘In this paper,a photoelectric device is introduced,which is used in detecting the quality on internal surface of thin and long steel pipe.In this device,the CCTV lens is used for extract tile flaw information on internal surface of the pipe,and make IBM-PC/AT 486 computer as controlling and image processing system.By this instrument,the functions,such as the digital conversion of input information,image processing,classification of recognition and output display can be obtained.In the petroleum and chemical industry,by using this apparatus,we can detect the quality on internal surface of various metal pipes with real-time automatically.
