摘要
高频直缝焊管的性能很大程度上取决于焊缝的质量,焊缝表面的质量成为了判别焊缝质量的重要因素。为了定量评价高频直缝焊管生产过程中焊缝表面形质量,采集了高频直缝焊管不同质量评级对应焊缝表面的图像信息,基于MATLAB平台,利用图像处理方法对不同质量焊缝表面图像进行了预处理,提取样本图像的几何形状及纹理特征等11维图像特征参数,经过对这些参数的动态变化建立了用于焊缝表面质量判别的BP神经网络模型。结果表明:缺陷边界周长﹑圆形度﹑矩形度等焊缝表面形貌图像特征参数可以作为描述焊缝表面形貌质量的定量指标,并可用BP网络模型对焊缝表面形貌进行识别预测。本研究为建立焊缝表面形貌的定量评价体系提供了新途径。
The performance of high frequency longitudinally welding pipe are depended on weld quality. The weld surface quality has became the discriminant of weld qualities in important factors. Quantitative monitoring of the morphological changes in the process of weld surface was researched in this paper. Based on MATLAB tool, it used the method of image processing to make pretreatment on weld surface morphology in different qualities, got the 11 dimensional features of geometry, texture characteristics. BP neural network model used to recognize the weld surface states was also established after parameters simplification. The results show that： the image features, such as Boundary perimeter, circularity, texture entropy ,rectangle degree can be used as quantitative parameters to describe the surface morphology of weld. It＇s also useful to use BP network model to recognize and judge surface morphology of weld. This study will supply a new approach to establish the quantitative evaluation system for the weld pipe production.
出处
《机械设计与制造》
北大核心
2017年第6期218-220,224,共4页
Machinery Design & Manufacture
关键词
焊缝质量
表面形貌
图像特征
分形特征
Weld Quality
Surface Morphology
Image Feature
Fractal Characteristics
