摘要
结合小波变换理论和Fourier相位相关算法,提出了一种亚像素配准算法。通过对图像进行小波分解获得低频系数,对低频系数应用相位相关进行粗配准,追踪到原图像,在原图像上进行精确配准。实验结果表明,该算法具有精度高,速度快,抗噪性能良好等优点。
A novel sub-pixel registration algorithm is proposed by combining wavelet transform with Fourier phase correlation algorithm.Wavelet decomposition of the images is performed and low frequency wavelet coefficients are obtained.Rough registration is performed in low frequency coefficients by phase correlation algorithm.The rough registration results are traced to the original image and a precise registration is performed in the original image.Experiment results show that the proposed algorithm has the advantages of high precision,less computing cost and good stability against noise.
出处
《计算机工程与应用》
CSCD
北大核心
2011年第7期187-188,207,共3页
Computer Engineering and Applications
基金
国家自然科学基金No.60672135~~
关键词
图像配准
亚像素
小波变换
相位相关
image registration
sub-pixel
wavelet transform
phase correlation
