希尔伯特变换
解析信号
傅里叶变换
希尔伯特谱分析
S变换
分数阶傅立叶变换
哈特利变换
数学
正交(天文学)
操作员(生物学)
梅林变换
数学分析
光学
物理
傅里叶分析
信号处理
小波变换
计算机科学
人工智能
光谱密度
小波
离散小波变换
统计
抑制因子
化学
电信
生物化学
转录因子
雷达
基因
作者
Kieran G. Larkin,Donald J. Bone,Michael A. Oldfield
标识
DOI:10.1364/josaa.18.001862
摘要
It is widely believed, in the areas of optics, image analysis, and visual perception, that the Hilbert transform does not extend naturally and isotropically beyond one dimension. In some areas of image analysis, this belief has restricted the application of the analytic signal concept to multiple dimensions. We show that, contrary to this view, there is a natural, isotropic, and elegant extension. We develop a novel two-dimensional transform in terms of two multiplicative operators: a spiral phase spectral (Fourier) operator and an orientational phase spatial operator. Combining the two operators results in a meaningful two-dimensional quadrature (or Hilbert) transform. The new transform is applied to the problem of closed fringe pattern demodulation in two dimensions, resulting in a direct solution. The new transform has connections with the Riesz transform of classical harmonic analysis. We consider these connections, as well as others such as the propagation of optical phase singularities and the reconstruction of geomagnetic fields.
科研通智能强力驱动
Strongly Powered by AbleSci AI