针对钢管缺陷的X射线检测方法,目前主要采用对数字式X射线(DR)影像进行人工判读的方式来检测缺陷,该方法受主观因素影响大,工作强度高,效率低。为替代人工检视方法,依据DR影像特征,研究了基于数字图像处理的缺陷检测算法。实验结果表明,采用的算法具有良好的适应性,比传统方法检测出的细小缺陷的效率更高,可满足对缺陷进行在线监测的需求。 At present, the defect of X-ray detection method for defects of steel pipe is detected by artificial interpretation of digital X-ray (DR) images. The method is greatly influenced by subjective factors, has high work intensity and low efficiency. In order to replace the manual inspection method, the defect detection algorithm based on digital image processing is studied based on DR image features. The experimental results show that the proposed algorithm has good adaptability and is more efficient than the small defects detected by the traditional method, which can meet the need of on-line monitoring of defects.