摘要
在工程机械中,很多机械关键部件为钢焊接结构。随着计算机视觉技术的发展,针对钢焊接结构容易出现裂纹、漏焊及焊缝外观不规则等缺陷检测效果差等问题,提出了基于微分形态学梯度处理的焊缝边缘检测方法。该算法首先采用中值滤波、白平衡处理和图像归一化等图像预处理技术纠正采集图像,然后用微分形态学梯度提取焊缝的二值化图像。仿真实验中使用微分形态学梯度算法与三种常用的图像处理算法对焊缝图像处理进行了对比与分析。结果表明,微分形态学梯度算法在处理焊缝图像时能够有效降噪,信噪比大且图像有用焊缝信息不丢失,对提高焊缝检测精度、保证机械机构安全具有重要意义。
In the field of construction machinery,many mechanical key components are steel welded structure.With the development of computer vision technology,a weld edge detection method based on differential morphological gradient processing is proposed to solve the problems of cracks,missing welding and irregular weld appearance in steel welded structures.Firstly,image preprocessing techniques such as median filtering,white balance processing and image normalization are used to correct the acquired image;and then the binary image of weld seam is extracted by the differential morphological gradient processing algorithm.In the simulation experiment,the differential morphological gradient algorithm is compared with three commonly used image processing algorithms,these are,Sobel operator,Prewitt operator and Canny algorithm.The results show that the differential morphological gradient algorithm can effectively reduce noise,it has high signal-to-noise ratio and the useful weld information of image is not lost.It is of great significance to improve the weld detection accuracy and ensure the safety of mechanical mechanism.
作者
张俊男
王远涛
ZHANG Jun-nan;WANG Yuan-tao(Department of Materials Engineering,Liaoning Mechatronics College,Dandong Liaoning 118009,China)
出处
《机械研究与应用》
2022年第2期45-48,共4页
Mechanical Research & Application
基金
辽宁机电职业技术学院科研项目:采煤机滚动轴承状态检测关键技术的研究(编号:ky202106)
辽宁机电职业技术学院科研项目:基于(HNC-CS)数控铣床的大赛设备开发(编号:ky202109)。
关键词
焊接
形态学
图像处理
焊缝信息
焊缝辨识
welding
morphology
image processing
weld information
weld identification
