Optical flow
From Wikipedia, the free encyclopedia
Optical flow is a concept which is close to, but not identical with the motion of objects within a visual representation. The difference is that true physical motion in the world of objects is not always reflected in gray value or color changes in the corresponding images, and gray value changes are not always due to motion of objects. Typically the motion is represented as vectors originating or terminating at pixels in a digital image sequence.
The term optical flow denotes in fact merely a vector field defined across the image plane, but the term is often (incorrectly) also used to denote the process of estimating the optical flow from image data, or even a particular class of algorithms used for that same purpose.
Estimating the optical flow is useful in pattern recognition, computer vision, and other image processing applications. It is closely related to motion estimation and motion compensation. Often the term optical flow is used to describe a dense motion field with vectors at each pixel, as opposed to motion estimation or compensation which uses vectors for blocks of pixels, as in video compression methods such as MPEG. Some consider using optical flow for collision avoidance and altitude acquisition system for unmanned air vehicles (UAVs).
[edit] Some methods for determining optical flow
- Phase correlation (inverse of normalized cross-power spectrum)
- Block correlation (sum of absolute differences, normalized cross-correlation)
- Gradient constraint-based registration
- Lucas Kanade method
- Horn Schunck method
