1. Basic information of the project: a kind of welded pipe detection device and its quality inspection method can be used to detect the real-time situation and early warning of high-frequency welding of welded pipe in high-speed movement, which can avoid the visual blind area, fatigue and damage during manual inspection, reduce the labor intensity and safety risk of employees, and improve the detection objectivity, welding efficiency and yield of welded pipe welding It has important practical significance and application value. The system uses advanced machine vision imaging and artificial intelligence data analysis technology to monitor the welding quality of welded pipe in high-speed motion in real-time, and extract the characteristics of welding points, including the shape and width of fusion line, the shape of weld heat affected zone, the shape of metal streamline, etc., combined with computer software and hardware system integration, communication and early warning protection system To ensure the safe, fast, reliable and efficient operation of welded pipe welding. 2. Core technical indicators: the detection resolution is not less than 5mm * 5mm, and the diameter of welded pipe is greater than 200mm 3. Upstream and downstream situation of the industry: adopting the scheme based on machine vision control system, adopting self-developed PLC control system and visual inspection algorithm, can improve the detection accuracy of welded pipe and reduce the error rate and missed detection rate of manual detection. 4. Conditions required for technology transformation: to be negotiated according to the actual application situation