Equation 5.2-15 for G*(ju) is the result of the complex convolution of G(ju) and Px(ju); it is in the so-called Poisson sum form, which helps visualize the effects of n=-œ
n=—oo the ideal sampling process in the spatial frequency domain. Figure 5.2-2A illustrates the spectrum of an ideally sampled, one-dimensional intensity, g(x), which is bandwidth limited so that it has no spectral energy above uN = us/2 r/mm, the Nyquist frequency (Northrop 1997). When |G(ju)| contains significant spectral energy above the Nyquist frequency, a phenomenon known as aliasing occurs, shown schematically in Figure 5.2-2B. The overlap of the high-frequency spectral components of the baseband of |G*(ju)| with the high-frequency portions of the first harmonic terms of |G*(ju)| generate an aliased, high-frequency portion of the baseband of |G*(ju)| that represents unrecoverable or lost information. Note that the more the spectrum of |G(ju)| extends past the spatial Nyquist frequency, the more spectral energy in the baseband of |G*(ju)| is unrecoverable. One way to avoid aliasing, whether sampling in the time or space domains, is to precede the sampler with a low-pass filter that attenuates high frequency energy in |G(ju)| beyond the spatial Nyquist frequency, uN = n/X r/mm. Such a low-pass filter is appropriately called an antialiasing filter.
In the CE, each point in the sampling array (i.e., each ommatidium) has a built-in low-pass filter that attenuates high spatial frequencies of the object and decreases aliasing. This filter is, of course, the DSF. To be effective, S(uN)/S(0) < 0.01. For the Hill DSF, given above,
Was this article helpful?
This guide will help millions of people understand this condition so that they can take control of their lives and make informed decisions. The ebook covers information on a vast number of different types of neuropathy. In addition, it will be a useful resource for their families, caregivers, and health care providers.
Post a comment