pintraregionuniformity

Computes the goodness measure based on intra-region uniformity.

Synopsis

pintraregionuniformity [-m mask] [rg_in|-] [im_in|-]

Description

pintraregionuniformity computes the intra-region uniformity criterion as defined by M. Levine & A. Nazif^*. This criterion relies on the idea that a good segmentation has a strong intra region uniformity. Uniformity is calculated from the inner contrast of the regions.

This criterion is near 1 when regions are homogeneous and near 0 when regions are heterogeneous.

The intra-region uniformity is computed as follows:

	1 - sum_R [ sum_s  ( ims[s] - mean(ims) )² ]

where s is a site (pixel) of a region.

Caution: Regions with label=0 are not considered for computing.

Inputs

rg_in: a region map.
im_in: a gray level image.

Result

Returns the measure between [0..1].

(Use pstatus to get this value).

Examples

Computes the intra-region uniformity measure for a simple binarization segmentation process:

   pbinarization 80 1e30 tangram.pan i1.pan
   plabeling 8 i1.pan i2.pan
   pintraregionuniformity i2.pan tangram.pan
   pstatus

C++ prototype

Errc PIntraRegionUniformity( const Reg2d &rg_in, const Img2duc &im_in );

Version française

Calcul du critère de qualité basé sur une mesure d'uniformité des régions.

Reference

^*M. D. Levine and A. M. Nazif, "Dynamic measurement of computer generated image segmentations", IEEE Trans. PAMI, 7(2): 155-164, 1985.

Author: Régis Clouard