多工件拼接焊缝面结构光滑度的视觉检测技术研究

Research on Visual Inspection Technology for Structural Smoothness of Multi workpiece Splicing Weld Surface

摘要:

由于多工件拼接焊缝面结构光滑度检测过程中,受其拼接面复杂且图像存在噪声,导致细节特征不明显,从而使检测精度、效率降低。为此,提出多工件拼接焊缝面结构光滑度的视觉检测技术。采用字典学习方法消除焊缝面图像中存在的噪声；将其输入MRFENet网络,提取图像特征,实现焊缝面图像的增强处理；采用增量二维主成分分析提取焊缝面结构的光滑度特征,实现光滑度检测。实验结果表明,所提方法图像处理效果好、检测精度高、检测效率高。

Abstract:

In the process of structural smoothness detection of multi workpiece splicing weld surface, due to the complexity of the splicing surface and the noise in the image, the details are not obvious, which reduces the detection accuracy and efficiency. For this reason, a visual inspection technology for the structural smoothness of multi workpiece splicing weld surface is proposed. Using dictionary learning methods to eliminate noise in weld surface images; Input it into the MRFENet network, extract image features, and achieve enhanced processing of weld surface images; Using incremental two-dimensional principal component analysis to extract the smoothness features of the weld surface structure and achieve smoothness detection. The experimental results show that the proposed method has good image processing performance, high detection accuracy, and high detection efficiency.