点击次数：
影响因子：7.6
DOI码：10.1016/j.patcog.2025.112212
所属单位：北京航空航天大学
教研室：宇航学院
发表刊物：Pattern Recognition
刊物所在地：英国
项目来源：国家自然科学基金
关键字：3D orthogonal moment; 3D spectrum analysis; Filter response; Terrain matching
摘要：Aiming to improve the terrain matching performance, the spectral analysis on the typical 3D orthogonal moments (OMs) is accomplished for the first time. The typical 3D OMs include Zernike moments (ZM), orthogonal Fourier-Mellin moments (OFMM), fractional-order Jacobi-Fourier moments (FJFM), exponent-Fourier moments (EFM), generic polar complex exponent transform (GPCET), and Bessel-Fourier moments (BFM). The general expression of the typical 3D OMs in both the spatial and the frequency domains is derived firstly. Then, the terrain spectrum is analyzed by the Fourier central slicing method. Both the spherical harmonic and the radial spectra of the typical 3D OMs are investigated further. Based on the spectral analysis on the typical 3D OMs, the terrain matching algorithms are designed accordingly. Numerical experiments indicate that the matching performance of the algorithms coincides with the spectral analysis results of the typical 3D OMs. These new findings are the foundation for designing an effective terrain matching algorithm.
论文类型：期刊论文
一级学科：控制科学与工程
文献类型：期刊
卷号：171
页面范围：1-11
是否译文：否
发表时间：2025-07-26
收录刊物：SCI