perosion

Synopsis

Description

perosion erodes points of stronger contrast according to a structuring element.

Erosion corresponds to the operation: replaces the central pixel p by the minimum of its neighbors where the neighbors are specified by the structuring element.

   erosion(p) = MIN(neighbors(p)). 

The structuring element is specified by its number num_se and its halfsize.

For a binary image, erosion erodes white areas.

For the region maps, erosion adds pixels with label=0 (background) at the points of erosion.

For the color images, the lexicographic is used: initially by using band X, in the event of equality by using the band Y then band Z.

Parameters

• num_se specifies the type of the structuring element:

  case of 2D:

  • 0: diamond (4-connexity)
  • 1: square (8-connexity)
  • 2: disc
  • 3: horizontal line
  • 4: diagonal line of 135 degrees (\)
  • 5: vertical line
  • 6: diagonal line of 45 degrees (/)
  • 7: cross
  • 8: X
  • 9: cercle

  case of 3D:

  • 0: bipyramid (6-connexity)
  • 1: cube (26-connexity)
  • 2: sphere
  • 3: horizontal line on x coordinate
  • 4: horizontal line on y coordinate
  • 5: horizontal line on x coordinate
  • 6: diagonal line on x-y coordinates (\)
  • 7: diagonal line on x-z coordinates (\)
  • 8: diagonal line on y-z coordinates (\)
  • 9: diagonal line on x-y coordinates (/)
  • 10: diagonal line on x-z coordinates (/)
  • 11: diagonal line on y-z coordinates (/)
  • 12: cross in 3d
  
  (This parameter is ignored for 1D image)
• halfsize specifies the half-size of the structuring element. For example, a half-size of 1 for a square gives a structuring element of size 3x3.

Inputs

• im_in: an image (1D, 2D, 3D) or a region map.

Outputs

• im_out: an image (or a region map) of the same type as im_in.

Result

Returns SUCCESS or FAILURE.

Examples

• Performs a White Top Hat with small 17x17 square structuring element:

   
     perosion 1 8 tangram.pan i1.pan
     pditation 1 i1.pan i2.pan
     pdif i2.pan tangram.pan out.pan
  

See also

C++ prototype

Version française

Erosion des points de fort contraste d'une image.