Abstract: Addressing the existing shortcomings of application of the currently-used X-ray weld seam testing technology for the SAWL pipe，i.e.，strong subjective dependence and poor result stability，a technical optimization path taking digitalization and intelligentization as its core is proposed. The scheme builds a close-loop system that integrates high-definition digital imaging，real-time noise reduction for the dynamic image，intelligent detection and evaluation of the the image defect，and digital delivery of the information as a whole. The core innovation of the said system is that it adopts the humanmachine collaboration evaluation mechanism as combining the dynamic image intelligent detection model with the manual recognition so as to fundamentally avoid any interference caused by subjective fluctuation factors of the operator such as professional level，proficiency level and physical and mental state，etc. The implementation of the system provides strong technical support for realizing the full automation of the pipe weld seam X-ray testing process，and the high reliability and stability of the detection result.