CAJ | 학술논문

图像配准是现代图像处理技术中的一项关键技术,在许多实际的应用领域都占有举足轻重的地位。文中介绍了一种结合尺度相互作用模型下墨西哥帽小波和归一化伪Zernike矩的图像配准方法。首先利用尺度相互作用模型下加入尺度因子的墨西哥帽小波分别提取参考图像和实时图像中的特征点,然后利用归一化伪Zernike矩不变量的方法和双向匹配策略对参考图像和实时图像的特征点进行匹配,用迭代加权最小二乘法估算出最佳仿射变换参数,最后用所得变换参数对实时图像进行变换和重采样来实现图像配准。实验结果表明：该算法能够精确提取和匹配特征点,有效地消除误匹配点对,被测加噪实物图像的特征点均方根误差为0.41,达到了像素级配准精度。
도상배준시현대도상처리기술중적일항관건기술,재허다실제적응용영역도점유거족경중적지위。문중개소료일충결합척도상호작용모형하묵서가모소파화귀일화위Zernike구적도상배준방법。수선이용척도상호작용모형하가입척도인자적묵서가모소파분별제취삼고도상화실시도상중적특정점,연후이용귀일화위Zernike구불변량적방법화쌍향필배책략대삼고도상화실시도상적특정점진행필배,용질대가권최소이승법고산출최가방사변환삼수,최후용소득변환삼수대실시도상진행변환화중채양래실현도상배준。실험결과표명：해산법능구정학제취화필배특정점,유효지소제오필배점대,피측가조실물도상적특정점균방근오차위0.41,체도료상소급배준정도。
Image registration is a key technique in modern image processing,and it is very important in many real applications. A method for image registration combining scale-interaction of Mexican-hat wavelets and normalized Pseudo-Zernike moments is proposed. First, feature points are extracted using scale-interaction of Mexican-hat wavelets in the reference image and sensed image respectively. Then, normalized Pseudo-Zernike moments and a bidirectional matching strategy are used to match them,and iterative weighted least square method is used to estimate the best affine transform parameters. At last,the sensed image is transformed and resampled to accomplish the image registration. The experiment indicates that the proposed algorithm extracts feature points and matches them exactly and eliminates wrong matched points effectively. The RMSE of the feature points of images of practicality with Gaussian noise is 0. 41 and it achieves pixel precision registration result.