摘要
为了实现对天基时间序列图像不同帧图像间坐标系的统一,设计了基于分块与互信息的天基序列图像的配准算法。首先,利用互信息对分块后的星图进行比对,对互信息最丰富的区域进行配准。接着,以尺寸不变特征转换算法,对图像进行特征点提取。最后在单应性变换矩阵的基础上,对配准图像进行配准计算得出配准系数。实验结果表明,配准算法时效性明显优于传统算法,精度略低于传统算法,满足后续目标检测需要。基本满足天基图像配准的时效性强、精度高、抗干扰能力强等要求。
To achieve the unification of coordinate systems between different frames of space-based time-series images,a registration algorithm for space-based sequence images based on blocking and mutual information is de-signed.First,mutual information is used to compare the blocked star maps,and registration is performed on the re-gion with the richest mutual information.Then,the Scale-Invariant Feature Transform algorithm is used to extract feature points from the images.Finally,based on the homography transformation matrix,registration calculations are performed on the registered images to obtain registration coefficients.The experimental results show that the timeliness of the registration algorithm is significantly better than that of traditional algorithms,and its accuracy is slightly lower than that of traditional algorithms,which meets the needs of subsequent target detection.It basically meets the re-quirements of space-based image registration,such as strong timeliness,high accuracy,and strong anti-interference ability.
作者
王一霖
颜昌翔
WANG Yi-lin;YAN Chang-xiang(Changchun Institute of Optics,Fine Mechanics,and Physics,Chinese Academy of Sciences,Changchun Jilin 130033,China;University of Chinese Academe of Sciences,Beijing 100049,China)
出处
《计算机仿真》
2025年第10期447-450,共4页
Computer Simulation
