摘要
针对印刷品质量检测中图像配准问题,提出了一种基于ROI(region of interest)模板的印刷图像配准算法.该方法以Canny算子获取的图像梯度向量为基础,定义相似性度量,同时结合阈值终止条件和矩阵搜索策略提高算法效率,并利用最小二乘法调整配准参数,最终实现彩色印刷图像高效配准.试验结果表明,相对于传统的图像配准算法,该方法配准精度高、速度快,而且鲁棒性强,能够为印刷品质量检测提供一种高效的图像配准算法.
A printing image registration algorithm based on ROI （region of interest） template is proposed for the problem of image registration in the quality inspection of printed matter. The similarity measurement is defined based on the image gradient vector obtained by Canny operator, which combines with threshold termination condition and matrix search strategy to improve algorithm efficiency. The registration parameters are adjusted by the least square method, and finally the color printing image registration is realized. Experimental results show that the proposed algorithm has higher accuracy, faster speed and stronger robustness than traditional template matching algorithm, which can provide an efficient image registration algorithm for printing quality inspection.
出处
《东华大学学报（自然科学版）》
CAS
CSCD
北大核心
2016年第4期582-586,592,共6页
Journal of Donghua University(Natural Science)
