胶印质量自动检测中图像配准的研究

胶印质量自动检测系统中,图像的配准是整个系统的关键技术。本实验提出使用SUSAN角点定位和模板匹配相结合的方法解决胶印质量自动检测中图像配准的问题。该方法首先利用SUSAN算法检测到图像角点,然后对图像进行平移和旋转,再在特定图像区域内,利用模板匹配的方法,进行精确配准。这种新方法与传统算法相比,能在提高定位精度的前提下,有效节省运算时间。实验结果表明新算法是有效且可行的。

Study on Image Registration in Offset Printing Quality Automatic Detection

In the system of offset quality automatic detection,image registration is a key technology.It can be resolved by the image registration algorithm combining the SUSAN algorithm with the template matching in the offset quality automatic detection.First,this algorithm used SUSAN algorithm for finding image corners,and the image was translated and rotated. Then,it was registered accurately by the template matching in the special region.Comparing with traditional algorithms,the operational time was saved effectively based on improving the registrable precision.Experiment results indicated that this algorithm is effective and feasible.