机器零件内部材质缺陷的无损探伤与自动检测

针对机器零部件内部缺陷与材质难于进行批量的自动定量检测,基于机器视觉,探讨了一种基于图像边缘提取的自动检测方法。文中剖析了工件探伤的检测机理,首先借助工业CT获取机器零部件图像,然后采用Canny算子提取图像的边缘信息,最后通过对图像的软件处理,自动地测量机器零部件的内部材质结构尺寸和缺陷尺寸。工业CT图像的实际测试数据验证了该探伤检测方法的精确度高与重复性好。实验结果表明,采用基于机器视觉的机器零部件自动探伤检测是可行与合理有效的。

Nondestructive Inspection and Automatic Detection of Internal Material Defect for Machine Parts

Aimming at being difficult to make automatic quantitative detection in batch size for internal de-fect and material quality of the machine parts,based on machine vision,the paper explored a sort of auto-matic detection method based on image edge extraction.In the paper,it made the anatomy of detecting mechanism for machine part inspection,first gained the machine part image by means of industrial CT, then extracted the image edge information by using the Canny operator,and finally measured the internal material structure size and the defect size of machine parts through the software processing for image auto-matically.It demonstrated that the method of flaw detection would be high in accuracy and good in repeat-ability for the actual test data of industrial CT image in machine parts.The experimental result shows that it is feasible,reasonable and effective to adopt the automatic flaw detection method based on machine vision.