Abstract: An X-ray remote real-time monitoring system using small-caliber tube was built up which can visually detect the weld porosity, slag, lack of fusion, incomplete penetration, cracks and other defects and also can provide defect alarm function. So developed system can overcome the effect of the operators subjective experience and technical limitations on the assessment of the quality of the films, and at the same time it can overcome the low detection efficiency, operation complexity and difficulty in realizing the automation of the detection process. An expert system based on the principles of reasoning and fuzzy neural network technology was built up for automatic tube weld defects identification, defect statistics, and issuing the relevant NDT records and reports. The system can provide the testing process of remote monitoring and real-time viewing product quality, thus improving the management level.