Abstract: In order to quickly detect the outer diameter and wall thickness of pipes, a collection and detection system based on image recognition was designed in this paper, which mainly included a mechanical structure part, an image acquisition part and an image processing part. The system can clamp and shoot the cross-section of pipes, and send the captured image to the PC to quickly detect the pipe diameter and wall thickness of the tested pipe. The image recognition system then uses Canny edge detection and contour feature extraction algorithm for identifying the pipe contour, and calculate the pipe diameter and wall thickness using scale of the image. Finally, experiments of pipes with different diameter, wall thickness and process methods were conducted. The results verified the high reliability and accuracy of the proposed system.