CAJ | 학술논문

针对 Keren 配准算法及其缺陷，提出了一种的 Keren 改进配准算法。该算法基于简化的四参数仿射变换模型而非传统的刚体变换模型，克服了传统 Keren 配准算法基于小角度旋转的局限，使得配准结果更加精确。仿真实验采用标准图像和车辆图像进行实验，结果表明本文提出的算法在大角度旋转下，能够获得比传统 Keren 配准算法更加精确的配准结果。
침대 Keren 배준산법급기결함，제출료일충적 Keren 개진배준산법。해산법기우간화적사삼수방사변환모형이비전통적강체변환모형，극복료전통 Keren 배준산법기우소각도선전적국한，사득배준결과경가정학。방진실험채용표준도상화차량도상진행실험，결과표명본문제출적산법재대각도선전하，능구획득비전통 Keren 배준산법경가정학적배준결과。
This paper introduces the Keren registration method in detail, and points out its disadvantage. Based on this disadvantage,this paper proposed an improvement Keren registration method. This improvement method bases on the four parameters affine transformation model rather than the rigid body transformation model. This method overcomes the limit the classic Keren registration method has, that it is only suitable for small rotation angel, and makes the registration more accuracy. The experiment is based on standard image and vehicle image, the result shows that the proposed method makes the registration result more accuracy than classic Keren registration method in case of large rotation angles.