Instantaneous phase

From Wikipedia, the free encyclopedia

In signal processing, a general sinusoidal signal with constant amplitude can be defined as:

s(t) = A\cdot \cos [\phi(t)] \

where A \ is the amplitude, and \phi(t)\, is the instantaneous phase (or local phase or simply phase) . The simplest useful form is:

\phi(t) = \omega t + \theta \,

which is effectively the same as the cyclical form:

(\omega t + \theta) \mod \ 2\pi \,,

where mod is the Modulo_operation.   \omega \, and \theta \, are constants.   \omega \, is an angular frequency (radians per second), which is related to ordinary frequency, f\, (in hertz) by: \omega = 2\pi f\,. Obviously, the frequency value determines the rate at which the phase changes. Therefore, it can be determined from the time derivative of the instantaneous phase, which in this case happens to be constant. But other forms of \phi(t)\, produce more general behavior. So the instantaneous angular frequency is defined as:

\omega(t) = \phi^\prime(t) = {d \over dt} \phi(t)\,

and the instantaneous frequency (Hz) is:

f(t) = \frac{1}{2 \pi} \phi^\prime(t) \.

Here we have also assumed the non-cyclical form of \phi(t)\,, whose continuity is not interrupted by the mod operator. That requires \phi(t)\, not be restricted to an interval of length . This unrestricted phase is sometimes referred to as unwrapped phase. Accordingly, the cyclical form is referred to as wrapped phase.

With the form:

\phi(t) = (\omega t + \theta) \mod \ 2\pi \,,

it is the case that:

  • φ(t) = 0 when s(t) assumes a local maximum value
  • φ(t) = π when s(t) assumes a local minimum value
  • \phi(t) = \pm {\pi \over 2} for all t where s(t) changes with maximum rate from a minimum to maximum value, or vice versa.

Consequently, for signals that are approximately sinusoidal, the phase value indicates local properties of the signal in terms of extreme points and sudden transitions from a larger value to a smaller, or vice versa. This can be used, e.g., in image processing and computer vision to detect points which are close to edges or lines, and also to measure the position of these points with sub-pixel accuracy.

  • Leon Coen, Time-Frequency Analysis, Prentice Hall, 1995.
  • Granlund and Knutsson, Signal Processing for Computer Vision, Kluwer Academic Publishers, 1995.

Advanced Search
Included Web Search Engines


Safe Search

close

Top Matching Results

Occasionally Search.com will highlight specialized results that are based on the context of your query. Examples of specialized results include specific links to news, images, or video.

Top Matching Results may highlight information from other Search.com pages, content from the CNET Network of sites, or third party content. The listings are based purely on relevance. Search.com does not receive payment for listings in this section but our partners that provide this data may get paid for listing these products.

Sponsored Links

This section contains paid listings which have been purchased by companies that want to have their sites appear for specific search terms and related content. These listings are administered, sorted and maintained by a third party and are not endorsed by Search.com.

Search Results

Search.com sends your search query to several search engines at one time and integrates the results into one list which has been sorted by relevance using Search.com's proprietary algorithm. You can customize the list of search engines included in your metasearch from the preferences.

The search engines that are used in your metasearch may allow companies to pay to have their Web sites included within the results. To view the Paid Inclusion policy for a specific search engine, please visit their Web site. Search.com does not accept payment or share revenue with any search engine partner for listings in this section.