The alignment of images greatly affects mosaic precision and components’quality in manufacturing of surface mounting electronic products. To acquire higher mosaic precision and better components’quality,a sub-pixel based algorithm was proposed to align the markers of local images,and to match and align the mosaiced image with the standard image. The matching and alignment error analyses for four common markers (circle,ellipse,square and rectangle)show for surface mounting components a rectangle or square marker is recommended,alignment error can roughly be 0. 2 pixel with the marker’s center as the alignment point and the marked data being processed through sub-pixel method. Furthermore,the sub-pixel based algorithm can acquire higher precision un-der fixed annoyance.% 在表面贴装电子组件的生产过程中,图像的匹配对准关系到图像拼接的精度与表面贴装电子组件的质量.为了提高图像的拼接精度与表面贴装电子组件的质量,提出一种基于亚像素的方法,对局部图像的标记进行匹配对准,对拼接后图像与标准图像进行匹配对准.经过对圆、椭圆、正方形和长方形等4种常用的标记进行匹配对准误差分析,结果表明：宜采用正方形或长方形作为表面贴装电子组件的标记,通过亚像素方法对标记数据进行处理,把标记中心点作为对准点,可以使对准误差减小到约0.2 pixel,且当噪声一定的情况下,采用亚像素方法精度更高.
展开▼
