上海交通大学学报(自然版) ›› 2017, Vol. 51 ›› Issue (6): 665-671.

• 兵器工业 • 上一篇    下一篇

 基于改进Sobel算法的焊缝X射线图像
气孔识别方法

 蒋华军a,蔡艳a,b,李超豪a,李芳a,b,华学明a,b   

  1.  上海交通大学 a. 上海市激光制造与材料改性重点实验室;
    b. 高新船舶与深海开发装备协同创新中心, 上海  200240
  • 出版日期:2017-06-30 发布日期:2017-06-30
  • 基金资助:
     

 Recognition Method for Gas Pores on XRay Image of Lap Joints
Based on the Improved Sobel Algorithm

 JIANG Huajuna,CAI Yana,b,LI Chaohaoa,LI Fanga,b,HUA Xueminga,b   

  1.  a. Shanghai Key Laboratory of Materials Laser Processingand Modification;
    b. Collaborative Innovation Center for Advanced Ship and DeepSea Exploration,
    Shanghai Jiao Tong University, Shanghai 200240, China
  • Online:2017-06-30 Published:2017-06-30
  • Supported by:
     

摘要:  提出一种具有自适应能力的焊缝X射线图像气孔检测方法.通过局部动态阈值分割法设计了多方向的焊缝X射线图像模板,采用Sobel算法对复杂背景下的焊缝进行边缘检测和区域标记,并在标记区域对原始图像x和y方向的灰度梯度进行分析,以增强小气孔和粘连气孔的识别能力.检测应用效果表明,所提出的方法具有良好的自适应能力且准确快速,能够有效地克服人工评片中产生的漏判与误判等缺点,并且能够存储和查询检测数据.

关键词:  , 焊缝, X射线图像, 气孔, 动态阈值, 灰度梯度

Abstract:   An adaptive recognition method for inner pores of welded joint was proposed based on Xray image. The Xray image templates were designed and established through local dynamic threshold method. The weld edge was extracted and marked from complicated background through improved Sobel algorithm. The gray gradients of original image in both x and y directions were investigated in the marked zone. The recognition capability for small pores and adhesive pores was enhanced. Application results showed that the proposed method had satisfied recognition capability, detection precision and response speed. It could effectively solve the problems of manual detection and provide platform for data storage and query.

Key words:  welded joint, Xray image, gas pores, dynamic threshold, gray gradient

中图分类号: