摘要
针对机械零件X射线图像的特点,对其缺陷的提取技术进行了研究,提出了基于迭代的阈值构造方法和基于数学形态学的边缘提取算法。针对在工作现场采集的射线探伤底片进行数字成像,结合图像的变换、中值滤波、阈值分割、边缘检测、特征提取、模式识别等算法,通过适当的处理和分析,对压力容器焊接缺陷进行自动检测与识别,最终实现了对焊接缺陷的机器自动识别。实践证明,该方法具有一定的快速性和实用性。
Defect extraction techniques are studied regarding the characteristic of X-ray images of spare parts ,and threshold-construction method based on iteration and edge-extraction algorithm based on mathematicalmorphology are advanced.A im at is work the spot collect to shoot line to explore wound film to carry on numeral be like ,combine the transformation of picture, in the value filter the examination,characteristic of the partition,edge of the wave ,Yu value to withdraw,the mode identify etc. calculate way,pass appropriate processing and analysis ,weld blemish to carry on an automatic examination and identify to the pressure container,feature extraction and pattern recognition and so forth .The experiments have shown thathis system is fast and practical.
出处
《机械设计与制造》
北大核心
2009年第2期78-80,共3页
Machinery Design & Manufacture
关键词
焊接缺陷
射线探伤
图像识别
Weld defect
X-ray inspection
Image recognition
