Image segmentation: Difference between revisions

Revision as of 16:56, 8 July 2004

Revision as of 05:52, 16 December 2004

Line 1:

In [[image analysis]], '''segmentation''' is the [[partition]] of a [[digital image]] <i>I</i> into multiple regions ([[set]]s of [[pixel]]s), according to some criterion.

The goal of segmentation is typically to locate certain ''objects of interest'' which may be depicted in the image. Segmentation could therefore be seen as a [[computer vision]] problem. However, many important segmentation algorithms are too simple to solve this problem accurately: thisthey limitationcompensate isfor compensatedthis limitation bywith their predictability, generality, and efficiency.

A simple example of segmentation is [[thresholding]] a [[grayscale]] image with a fixed threshold <i>''t</i>'': each pixel <i>''p</i>'' is assigned to one of two classes, <i>''P</i>₀'' or <i>''P</i>₁'', depending on whether <i>''I</i>(<i>p</I>) &lt; <i>t</i>'' or <i>''I</i>(<i>p</I>) &ge; <i>t</i>''.

Some other segmentation algorithms are based on segmenting images into regions of similar texture according to [[Wavelet|wavelet]] or [[Fourier transform|Fourier]] transforms.

Segmentation criteria can be arbitrarily complex, and take into account global as well as local criteria. A common requirement is that each region must be [[connected]] in some sense.