基于序贯相似性测算法的图像模板配准算法

A Fast Image Template Registration Algorithm Based on SSDA

图像配准是很多图像处理应用中非常重要的一个环节.针对烟标印刷缺陷检测系统的模板配准算法在此得到了较深入的研究.在传统的序贯相似性检测算法(SSDA)基础上,提出一种新的配准算法.实验证明,该算法比传统的SSDA算法在速度上有较大提高,能满足在线烟标检测系统中图像处理算法的实时性要求.

Image registration is an important component in many image processing applications. This article studies the template matching algorithm for the cigarette label printing defects inspection system. Based on the traditional Sequential Similarity Detection Algorithm （SSDA）, a new matching algorithm is proposed. Experimental results indicate that the suggested algorithm obtains a higher speed compared with the traditional SSDA, and it meets the speed requirement of the real-time image processing algorithms for the on-line defect inspection on cigarette labels.