摘要
工业产品外观的瑕疵会直接造成产品商业价值的贬值。一般的视觉检测方法是利用环形光、同轴光等光源,结合单目视觉进行缺陷检测,但是对于一些存在的凸起、凹陷等具有三维特征的缺陷,由于其表面光学反射特性不明显,很难获取到缺陷的信息。结合三维视觉检测技术提出基于光栅投影的表面凹凸缺陷检测方法,投射棋盘格图案进行定位和角点提取,从中可以获得稀疏的点云数据,完成对凹坑缺陷的实时检测。
Defects in the appearance of industrial products will directly cause the depreciation of the commercial value of the product.The general visual inspection method is to use ring light,coaxial light and other light sources,combined with monocular vision for defect detection.However,for some defects with three-dimensional characteristics such as protrusions and depressions,the optical reflection characteristics of the surface are not obvious.It is difficult to obtain information about defects.Combining the three-dimensional visual inspection technology,this paper proposes a surface bump defect detection method based on raster projection,projecting a checkerboard pattern for positioning and corner extraction,from which sparse point cloud data can be obtained to complete the real-time detection of pit defects.
出处
《工业控制计算机》
2021年第11期105-107,110,共4页
Industrial Control Computer
关键词
凹凸缺陷检测
光栅投影
角点提取
实时
bump defect detection
raster projection
corner point extraction
real-time
